--------------------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  /Users/erzoluttmer/Dropbox/AnnuityVignettes/VignFINAL_REStat/Analysis/annuityanalyses.log
  log type:  text
 opened on:  13 Dec 2019, 11:00:56

. clear

. set linesize 140

. 
. // ******************************  AnnuityAnalyses.do  ***********************************
. **
. ** This program performs all the analyses for the paper:
. ** 
. **      Behavioral Impediments to Valuing Annuities: 
. **            Complexity and Choice Bracketing
. ** 
. ** to be published in the Review of Economics and Statistics in 2020
. ** 
. **                           by
. **       Jeffrey R. Brown, Arie Kapteyn, Erzo F.P. Luttmer, 
. **             Olivia S. Mitchell, and Anya Samek
. **
. ** This program uses two datasets:
. ** 1. UAS49_WithoutID -- Data from wave 49 of the Understanding America Study (UAS)
. **                       augmented with select variables from other waves of 
. **                       of the UAS. Variables such as UasID, state of residence,
. **                       and other variables that could help identify the respondents
. **                       are suppressed.
. **
. **                       IMPORTANT: If you use this UAS data, kindly email your 
. **                                  full name and affiliation to uas-l@usc.edu 
. **                                  so that the administrators of the Understanding 
. **                                  America Study can keep track of those who use 
. **                                  their data.
. **
. **                       You can access the same data but including identifiers
. **                       (which allows you to merge the data to any other wave in
. **                       the UAS) by setting up an account with the UAS at
. **                       uasdata.usc.edu
. **
. ** 2. cps_ASEC2016.dat - CPS abstract with basic demographics obtained from 
. **                       the IPUMScps (http://cps.ipums.org/cps/)
. **
. **                       Year: 2016 Annual Social and Economic Supplement (2016 used to match 
. **                            timing of UAS sample)
. **                       Age selection: 18+
. **
. **
. **
. ** This program produces the following output:
. **
. ** 1. AnnuityAnalyses.log -- Log file with the output for all the tables.
. **
. **                    To search for a table, search for "Table #" for 
. **                         normal tables or "Table A##" for appendix tables
. **                    To search for a figure, search for "Figure #" for 
. **                         normal figures or "Figure A#" for appendix figures
. **                    To serach for a text claim, search "Text Claim ##" 
. **                         where ## is the determined by the order in which they 
. **                         appear in the text, or alternatively, simply search on
. **                         a snipped of the text in the paper than contains an
. **                         empirical result that is not in the tables.
. **
. ** 2. Excel files with the data for the figures:
. **
. **                    rawdata_fig1a_midbuy.xls   - Figure 1, CDF of buy values
. **                    rawdata_fig1b_midsell.xls  - Figure 1, CDF of sell values
. **                    rawdata_fig2_logdiff.xls   - Figure 2, CDF of log(sell/buy)
. **                    rawdata_figA1a_midbuy.xls  - Appendix Figure A1, CDF of buy values
. **                    rawdata_figA1b_midsell.xls - Appendix Figure A1, CDF of sell values
. **                    rawdata_figA2_spread.xls   - Appendix Figure A2, CDF of spread
. **                 
. // ***************************************************************************************
. 
. 
. 
. 
. 
. 
. 
. // ***************************************************************************************
. // ***************************************************************************************
. //
. //  PART 1: Data cleaning and definitions of varialbes
. //
. // ***************************************************************************************
. // ***************************************************************************************
. 
. 
. 
. ** Data from Wave49 of the Understanding America Study
. **
. ** This wave contained our experiment.
. **
. **
. ** IMPORTANT: If you use this UAS data, kindly email your 
. **            full name and affiliation to uas-l@usc.edu 
. **            so that the administrators of the Understanding 
. **            America Study can keep track of those who use 
. **            their data.
. **
. use uas49_withoutid

. 
. 
. 
. 
. 
. // ***************************************************************************************
. //            Label, check, and name the primary treatment variables 
. // ***************************************************************************************
. 
. ** give variables that contain the randomizations more intuitive names
. ** -------------------------------------------------------------------
. rename randomizer_advice_ssb          ss_benefit

. rename randomizer_advice_lsstartvalue ls_startvalue

. rename randomizer_name                vignette_name

. rename randomizer_advice_intro        complexity

. rename ed_001                         test_question1

. rename ed_002                         test_question2

. 
. **** Create Binary Variables with "Yes" corresponding to 1 and "No" corresponding to 0
. **** with name that makes clear what 1 corresponds to
. **** All newly generated variables are also labeled and checked to ensure no missing values
. 
. ** define 0 as No and 1 as Yes
. label define noyes 0 "No" 1 "Yes"                       

. 
. ** 1 means person received consequence message treatment
. gen byte  consequence = 2 - randomizer_education                

. label var consequence "Consequence Treatment"

. label val consequence noyes

. assert    consequence < .

. drop randomizer_education                                               

. 
. ** 1 means Lump Sum amount was mentioned first
. gen byte  ls_first  = 2 - randomizer_advice_answer_order        

. label var ls_first "LS option mentioned first"

. label val ls_first noyes

. assert    ls_first < . 

. drop randomizer_advice_answer_order

. 
. ** 1 means person was reminded of quick spend down consequences first when receiving message
. ** It is defined for everyone but only affected those receiving the consequence message
. gen byte  quick_first = 2 - randomizer_education_block

. label var quick_first "Conseq. msg.: quick spend down mentioned first"

. label val quick_first noyes

. assert    quick_first < .

. drop randomizer_education_block

. 
. ** 1 means sell question was asked first (0 means buy question was asked first)
. gen byte  sell_first = 2 - randomizer_advice_order

. label var sell_first "Sell question first, then buy question"

. label val sell_first noyes

. assert    sell_first < .

. drop randomizer_advice_order

. 
. 
. **** Create dummies for each randomization variable that is not binary ****
. **** Also check for missing values and label newly created dummies     ****
. ** ------------------------------------------------------------------------
. 
. ** starting lump sum value proposed in exchange for annuity increase/decrease
. tab       ls_startvalue, gen(ls_startvalue_)

 indicates LS start |
   value R received |      Freq.     Percent        Cum.
--------------------+-----------------------------------
   1 LS Low: $10000 |      1,577       34.31       34.31
2 LS Medium: $20000 |      1,542       33.55       67.86
  3 LS High: $30000 |      1,477       32.14      100.00
--------------------+-----------------------------------
              Total |      4,596      100.00

. assert    ls_startvalue < .

. label var ls_startvalue_1 "LS low: 10k"         //starting LS value of $10,000

. label var ls_startvalue_2 "LS medium: 20k"      //starting LS value of $20,000

. label var ls_startvalue_3 "LS high: 30k"        //starting LS value of $30,000

. 
. ** name/gender assigned to primary vignette person
. ** (i.e., the persons to whom the respondent gives advice on buying/selling
. **        the annuity)
. tab       vignette_name, gen(vignette_name_)

   indicates |
  which name |
and gender R |
    received |      Freq.     Percent        Cum.
-------------+-----------------------------------
 1 Mr. Jones |      1,205       26.22       26.22
2 Mrs. Jones |      1,125       24.48       50.70
 3 Mr. Smith |      1,083       23.56       74.26
4 Mrs. Smith |      1,183       25.74      100.00
-------------+-----------------------------------
       Total |      4,596      100.00

. assert    vignette_name < .

. label var vignette_name_1 "Mr. Jones"

. label var vignette_name_2 "Mrs. Jones"

. label var vignette_name_3 "Mr. Smith"

. label var vignette_name_4 "Mrs. Smith"

. 
. ** type of complexity/no complexity added to vignette
. tab       complexity, gen(complexity_)

 indicates the intro R received for the |
                       advice questions |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
1 Narrow spread in age of death. No unc |      1,561       33.96       33.96
2 Wide spread in age of death. Uncertai |      1,487       32.35       66.32
3 Adding Additional Information to Add  |      1,548       33.68      100.00
----------------------------------------+-----------------------------------
                                  Total |      4,596      100.00

. assert    complexity < .

. label var complexity_1 "No complexity: Narrow Spread"

. label var complexity_2 "Complexity: Wide Spread"

. label var complexity_3 "Complexity: Added Info"

.  
. ** initial Social Security benefit of the vignette person
. tab       ss_benefit, gen(ss_benefit_)

  indicates |
 advice ssb |
 R received |      Freq.     Percent        Cum.
------------+-----------------------------------
      1 800 |      1,170       25.46       25.46
     2 1200 |      1,118       24.33       49.78
     3 1600 |      1,189       25.87       75.65
     4 2000 |      1,119       24.35      100.00
------------+-----------------------------------
      Total |      4,596      100.00

. assert    ss_benefit < .

. label var ss_benefit_1 "SSbenefit 800"          //initial SS benefit of $800/month

. label var ss_benefit_2 "SSbenefit 1200"         //initial SS benefit of $1,200/month

. label var ss_benefit_3 "SSbenefit 1600"         //initial SS benefit of $1,600/month

. label var ss_benefit_4 "SSbenefit 2000"         //initial SS benefit of $2,000/month

.  
. ** Also create Social Security benefits in hundreds of dollars
. gen       ss_benefit100dollar = 4 + ss_benefit*4

. label var ss_benefit100dollar "SS benefit (in $100)"

.  
.  
.  
. **********  Treatment variables derived from primary treatment variables        *************
. **********      these variables are also labeled, defined, and checked for missing values********
. ** ------------------------------------------------------------------------------------------
. 
.  
. ** Indicator for any complexity treatment (either wide spread or added info)
. gen       any_complexity = complexity==2 | complexity==3

. label var any_complexity "Any Complexity"

. assert    any_complexity < .

.  
. ** The key treatments interacted
. ** 2x3 design, meaning different types of complexity are examined individually
. gen     treat2x3 = 1 if complexity==1 & consequence==0  // no complexity          & no consequene msg
(3,810 missing values generated)

. replace treat2x3 = 2 if complexity==2 & consequence==0  // wide spread complexity & no consequence msg
(726 real changes made)

. replace treat2x3 = 3 if complexity==3 & consequence==0  // added info complexity  & no consequence msg
(757 real changes made)

. replace treat2x3 = 4 if complexity==1 & consequence==1  // no complexity          & yes consequence msg
(775 real changes made)

. replace treat2x3 = 5 if complexity==2 & consequence==1  // wide spread complexity & yes consequence msg
(761 real changes made)

. replace treat2x3 = 6 if complexity==3 & consequence==1  // added info complexity  & yes consequence msg
(791 real changes made)

. 
. ** Label and define each of the 2x3 treatment interactions
. label var treat2x3 "Main treatments, 2x3 design"

. label def treat2x3  1 "Baseline"  ///
>                     2 "Wide Spread; NO conseq. msg"     ///
>                     3 "Added Info ; NO conseq. msg"     ///
>                     4 "Narrow Spread; CONSEQ. msg"      ///
>                     5 "Wide Spread  ; CONSEQ. msg"      ///
>                     6 "Added Info   ; CONSEQ. msg"              

. label val treat2x3 treat2x3                    

. 
. ** Generate and label dummies for each of the 2x3 key treatments
. tab treat2x3, gen(treat2x3_)

Main treatments, 2x3 design |      Freq.     Percent        Cum.
----------------------------+-----------------------------------
                   Baseline |        786       17.10       17.10
Wide Spread; NO conseq. msg |        726       15.80       32.90
Added Info ; NO conseq. msg |        757       16.47       49.37
 Narrow Spread; CONSEQ. msg |        775       16.86       66.23
 Wide Spread  ; CONSEQ. msg |        761       16.56       82.79
 Added Info   ; CONSEQ. msg |        791       17.21      100.00
----------------------------+-----------------------------------
                      Total |      4,596      100.00

. label var treat2x3_1 "Baseline"  

. label var treat2x3_2 "Wide Spread; NO conseq. msg" 

. label var treat2x3_3 "Added Info ; NO conseq. msg" 

. label var treat2x3_4 "Narrow Spread; CONSEQ. msg"  

. label var treat2x3_5 "Wide Spread  ; CONSEQ. msg"  

. label var treat2x3_6 "Added Info   ; CONSEQ. msg"

. 
. 
. ** The key treatments interacted
. ** 2x2 design with any complexity instead of analyzing complexities separately
. gen     treat2x2 = 1 if any_complexity==0 & consequence==0      // no complexity  & no consequence msg
(3,810 missing values generated)

. replace treat2x2 = 2 if any_complexity==1 & consequence==0      // any complexity & no consequence msg
(1,483 real changes made)

. replace treat2x2 = 3 if any_complexity==0 & consequence==1      // no complexity  & yes consequence msg
(775 real changes made)

. replace treat2x2 = 4 if any_complexity==1 & consequence==1      // any complexity & yes consequence msg
(1,552 real changes made)

. 
. ** Label and define each of the 2x2 treatment interactions
. label var treat2x2 "Main treatments, 2x2 design"   

. label def treat2x2  1 "Baseline"       ///      
>                     2 "Complexity; NO conseq. msg"      ///
>                     3 "NO Complexity; CONSEQ. msg"  ///
>                     4 "Complexity   ; CONSEQ. msg"                   

. label val treat2x2 treat2x2   

. 
. ** Generate and label dummy variables for each of the 2x2 treatments                 
. tab treat2x2, gen(treat2x2_)

      Main treatments, 2x2 |
                    design |      Freq.     Percent        Cum.
---------------------------+-----------------------------------
                  Baseline |        786       17.10       17.10
Complexity; NO conseq. msg |      1,483       32.27       49.37
NO Complexity; CONSEQ. msg |        775       16.86       66.23
Complexity   ; CONSEQ. msg |      1,552       33.77      100.00
---------------------------+-----------------------------------
                     Total |      4,596      100.00

. label var treat2x2_1 "Baseline"                 

. label var treat2x2_2 "Complexity; NO conseq. msg" 

. label var treat2x2_3 "NO Complexity; CONSEQ. msg" 

. label var treat2x2_4 "Complexity   ; CONSEQ. msg"      

. 
. 
. 
. **************** Check and create the Sell / Buy variables ***********
. ** -------------------------------------------------------------------
. 
. ** Rename variables measuring whether buy or sell offer was accepted
. ** and label all sell and buy variables using a foreach loop
. foreach num of numlist 1/5{
  2.   gen byte sell`num' = ad_001_`num' - 1
  3.   drop ad_001_`num'
  4.   
.   gen byte buy`num' = ad_002_`num' - 1
  5.   drop ad_002_`num'
  6.   
.   label variable sell`num' "sell`num'"
  7.   label variable buy`num' "buy`num'"
  8. }
(26 missing values generated)
(28 missing values generated)
(27 missing values generated)
(30 missing values generated)
(29 missing values generated)
(32 missing values generated)
(30 missing values generated)
(34 missing values generated)
(32 missing values generated)
(38 missing values generated)

. 
. ** Label the sell and buy responses and the randomizations as noyes variables
. *1 meaning buy/sell was accepted, 0 meaning buy/sell was rejected
. label val sell1 sell2 sell3 sell4 sell5 buy1 buy2 buy3 buy4 buy5 noyes

. 
. ** convert the offered lumpsum amounts into numererics (removing commas)
. ** creates sellprice'num' and buyprice'num' variables using a foreach loop
. foreach num of numlist 1/5{
  2.   destring flsellpayment_`num'_, ignore(",") replace
  3.   gen sellprice`num'=flsellpayment_`num'_
  4.   label variable sellprice`num' "sellprice`num'"
  5.   
.   destring flmakepayment_`num'_, ignore(",") replace
  6.   gen buyprice`num'=flmakepayment_`num'_
  7.   label variable buyprice`num' "buyprice`num'"
  8. }
flsellpayment_1_: byte , removed; replaced as int
(23 missing values generated)
(23 missing values generated)
flmakepayment_1_: byte , removed; replaced as int
(22 missing values generated)
(22 missing values generated)
flsellpayment_2_: byte , removed; replaced as long
(26 missing values generated)
(26 missing values generated)
flmakepayment_2_: byte , removed; replaced as long
(28 missing values generated)
(28 missing values generated)
flsellpayment_3_: byte , removed; replaced as long
(27 missing values generated)
(27 missing values generated)
flmakepayment_3_: byte , removed; replaced as long
(30 missing values generated)
(30 missing values generated)
flsellpayment_4_: byte , removed; replaced as long
(29 missing values generated)
(29 missing values generated)
flmakepayment_4_: byte , removed; replaced as long
(32 missing values generated)
(32 missing values generated)
flsellpayment_5_: byte , removed; replaced as long
(30 missing values generated)
(30 missing values generated)
flmakepayment_5_: byte , removed; replaced as long
(34 missing values generated)
(34 missing values generated)

. 
. **** Check that the survey was implemented as we instructed ****
. ** 1. Check that starting value specified was used
. assert sellprice1==10000 if ls_startvalue==1 & sellprice1<.

. assert sellprice1==20000 if ls_startvalue==2 & sellprice1<.

. assert sellprice1==30000 if ls_startvalue==3 & sellprice1<.

. 
. ** 2. rejected sell prices should lead to question about a higher sell price next
. assert sellprice2 > sellprice1 if sell1==0

. assert sellprice3 > sellprice2 if sell2==0

. 
. * The survey instrument messed up for this one observation 
. * because after rejecting $100k, the person should have been asked 
. * about $200k, not about about 50k
. * Perhaps the use of the backbutton induced this.
. gen messedup = (sellprice4 <= sellprice3 & sell3==0)

. count if messedup
  1

. 
. * Same assertion as before but now we account for "messed up" observation
. assert sellprice4 > sellprice3 if sell3==0 & ~messedup  

. assert sellprice5 > sellprice4 if sell4==0 & ~messedup

. 
. ** 3. accepted sell price should lead to question about a lower sell price next
. assert sellprice2 < sellprice1 if sell1==1

. assert sellprice3 < sellprice2 if sell2==1

. assert sellprice4 < sellprice3 if sell3==1 

. assert sellprice5 < sellprice4 if sell4==1 

. 
. ** 4. rejected buy price should lead to question about a lower buy price next
. assert buyprice2 < buyprice1 if buy1==0

. assert buyprice3 < buyprice2 if buy2==0

. assert buyprice4 < buyprice3 if buy3==0 

. assert buyprice5 < buyprice4 if buy4==0 

. 
. ** 5. accepted buy price should lead to question about a higher buy price next
. assert buyprice2 > buyprice1 if buy1==1

. assert buyprice3 > buyprice2 if buy2==1

. assert buyprice4 > buyprice3 if buy3==1 

. assert buyprice5 > buyprice4 if buy4==1 

. 
. ** Determine the supremum and infimum of the buy and sell prices based on
. ** i)  (sup) the lowest sell price accepted and (inf) highest sell price rejected, and
. ** ii) (inf) the highest buy price accepted and (sup) the lowest buy price rejected.
. **
. ** Topcode the sell & buy price at $million, which is twice the highest sell or buy price offered,
. ** and bottom code the sell and buy price at $0.
. **
. ** This follows Brown, Kapteyn, Luttmer, Mitchell, JEEA 2017
. **
. ** The rationale of this topcode is that if the distribution of valuations
. ** is a Pareto distribution with parameter 2 (as seems to be the case for many 
. ** valuation, income, and wealth distributions in the right tail), then
. ** the expected value conditional on exceeding the topcode is twice the topcode.
. **
. ** Note: because of the somewhat arbrary nature of the topcode, the mean in levels
. ** is not that meaningful because it is sensitive to this topcode. 
. 
. * generate numbered variables for all of the accepted sell prices, 
. * rejected sell prices, accepted buy prices, and rejected buy prices 
. * using a foreach loop
. foreach num of numlist 1/5{
  2. gen sellpriceyes`num'=sellprice`num' if sell`num'==1
  3. gen  sellpriceno`num'=sellprice`num' if sell`num'==0
  4. gen  buypriceyes`num'= buyprice`num' if  buy`num'==1
  5. gen   buypriceno`num'= buyprice`num' if  buy`num'==0
  6. }
(1,926 missing values generated)
(2,696 missing values generated)
(3,456 missing values generated)
(1,168 missing values generated)
(2,572 missing values generated)
(2,051 missing values generated)
(2,756 missing values generated)
(1,870 missing values generated)
(2,309 missing values generated)
(2,316 missing values generated)
(2,926 missing values generated)
(1,702 missing values generated)
(2,451 missing values generated)
(2,175 missing values generated)
(2,764 missing values generated)
(1,866 missing values generated)
(2,694 missing values generated)
(1,934 missing values generated)
(2,576 missing values generated)
(2,058 missing values generated)

. 
. ** First code the sell price
. ** Find the sup and inf sell prices using rowmin and rowmax functions
. egen sellpricesup=rowmin(sellpriceyes1 sellpriceyes2 sellpriceyes3 sellpriceyes4 sellpriceyes5)
(374 missing values generated)

. egen sellpriceinf=rowmax(sellpriceno1  sellpriceno2  sellpriceno3  sellpriceno4  sellpriceno5) 
(277 missing values generated)

. 
. ** Label sup and inf sell variables
. label variable sellpricesup "upper bound on selling price"

. label variable sellpriceinf "lower bound on selling price"

. 
. ** implement top and bottom coding
. ** Topcode sell price at $1,000,000 - twice the highest sell price offered
. mvencode sellpricesup, mv(1000000)
sellpricesup: 374 missing values recoded

. 
. ** Bottomcode sell price at $0
. mvencode sellpriceinf, mv(0)
sellpriceinf: 277 missing values recoded

. 
. ** Create sellmissing variable to check for missing sell values
. ** If sell variable is missing or messed up, set sup and inf values to missing
. gen sellmissing = missing(sell1, sell2, sell3, sell4, sell5)

. replace sellpricesup = . if sellmissing | messedup
(33 real changes made, 33 to missing)

. replace sellpriceinf = . if sellmissing | messedup
(33 real changes made, 33 to missing)

. 
. ** Check to make sure there are no unintentional missing values
. assert  sellpricesup < . if ~(sellmissing | messedup)

. assert  sellpriceinf < . if ~(sellmissing | messedup)

. 
. ** Second, code the buy price
. ** Find the inf and sup buy prices using rowmax and rowmin functions
. egen buypriceinf=rowmax(buypriceyes1 buypriceyes2 buypriceyes3 buypriceyes4 buypriceyes5)
(911 missing values generated)

. egen buypricesup=rowmin(buypriceno1  buypriceno2  buypriceno3  buypriceno4  buypriceno5) 
(402 missing values generated)

. 
. ** Label sup and inf buy variables
. label var buypricesup "upper bound on buying price"

. label var buypriceinf "lower bound on buying price"

. 
. ** implement top and bottom coding same as for sell prices
. mvencode buypricesup, mv(1000000)
 buypricesup: 402 missing values recoded

. mvencode buypriceinf, mv(0)
 buypriceinf: 911 missing values recoded

. 
. ** Create buymissing to check for missing buy values
. ** If buy variable is missing or messed up, set sup and inf values to missing
. gen buymissing = missing(buy1, buy2, buy3, buy4, buy5)

. replace buypricesup = . if buymissing | messedup
(39 real changes made, 39 to missing)

. replace buypriceinf = . if buymissing | messedup
(39 real changes made, 39 to missing)

. 
. ** Check to make sure no unintentional missing values
. assert  buypricesup < . if ~(buymissing | messedup)

. assert  buypriceinf < . if ~(buymissing | messedup)

. 
. ** Define log-midpoints of annuity valuations
. ** and the spread variable, following the definition of BKLM(2017)
. * Generate and label log-midpoints and spread variables
. gen double midbuy         = (buypricesup +buypriceinf)/2        // avg of buy sup and inf
(39 missing values generated)

. gen double midsell        = (sellpricesup+sellpriceinf)/2       // avg of sell sup and inf
(33 missing values generated)

. gen double logbuyprice    = log(midbuy)                                         // log Buy                      
(39 missing values generated)

. gen double logsellprice   = log(midsell)                                        // log of midsell
(33 missing values generated)

. gen double logspread      = abs(logsellprice-logbuyprice)       // abs diff between logsell and logbuy
(44 missing values generated)

. gen double logsellbuydiff =     logsellprice-logbuyprice        // same as logspread without abs
(44 missing values generated)

. gen double meanlogprice   =    (logsellprice+logbuyprice)/2     // avg of logsell and logbuy price
(44 missing values generated)

. 
. label var midbuy         "Buy price (midpoint)"

. label var midsell        "Sell price (midpoint)"

. label var logbuyprice    "Log Buy"

. label var logsellprice   "Log Sell"

. label var logspread      "Log Spread" 

. label var logsellbuydiff "Log Sell - Log Buy"

. label var meanlogprice   "Mean of log sell and log buy price"

. 
. ** Given that the vignette person has $100k in savings, they cannot not logically spend more that 100k
. ** on the annuity. Hence there is a logical (though not mechanical) topcode on the buy value.
. ** To make sure this is not driving anything, we create a spread variable where we put a mechanical topcode on
. ** both the buy and the sell value
. gen logspread_100k = abs(log((min(100000,sellpricesup)+min(100000,sellpriceinf))/2)   ///
>                        - log((min(100000, buypricesup)+min(100000, buypriceinf))/2))  if logspread<.
(44 missing values generated)

. 
. 
. 
. 
. 
. ************ Generate and label the demographic control variables ***************
. ** ------------------------------------------------------------------------------
.         
. ** dividing the age^2 by 100 keeps coefficients legible
. gen agesq = age^2 / 100
(7 missing values generated)

. label var agesq "Age squared divided by 100"

.   
. ** age categories for tables with summary statistics
. recode age (18/34=1 "18-34") (35/49=2 "35-49") (50/64=3 "50-64") (65/106=4 "65+"), gen(agecat)
(4589 differences between age and agecat)

. gen agecat_18_34=agecat==1

. gen agecat_35_49=agecat==2

. gen agecat_50_64=agecat==3

. gen agecat_65_plus=agecat==4

. label var agecat "age categories"

. label var agecat_18_34      "Age 18-34"   

. label var agecat_35_49      "Age 35-49"   

. label var agecat_50_64      "Age 50-64"   

. label var agecat_65_plus    "Age 65+" 

. 
. 
.   
.   
. ** Generate female, married, and race/ethnicity categories
. ** Check to make sure no missing demographic values
. ** 1 means person is female, 
. gen       female = 1 - gender 
(3 missing values generated)

. label var female "Female"

. label val female noyes

. 
. ** 1 means person is married
. gen married =  marital==1 | marital==2 if marital < .
(3 missing values generated)

. label var married "Married"

. label val married noyes

. 
. ** 1 means person is non-hispanic white
. gen       nhwhite =  race==1 & hisplatino==0 if ~missing(race, hisplatino)
(17 missing values generated)

. label var nhwhite "Non-Hispanic White"

. label val nhwhite noyes

. 
. ** 1 means person is non-hispanic black
. gen       nhblack =  race==2 & hisplatino==0 if ~missing(race, hisplatino)
(17 missing values generated)

. label var nhblack "Non-Hispanic Black"

. label val nhblack noyes

. 
. ** 1 means person is hispanic and any race
. gen       hispanic = hisplatino==1           if ~missing(race, hisplatino)
(17 missing values generated)

. label var hispanic "Hispanic, Any Race"

. label val hispanic noyes

. 
. ** 1 means person is non-hispanic and not white or black
. gen       nhother = (race==3 | race==4 | race==5 | race==6) & hisplatino==0 if ~missing(race, hisplatino)
(17 missing values generated)

. label var nhother "Other Race/Ethnicity"

. label val nhother noyes

. 
. ** Check to make sure all respondents were placed into race/ethnicity groups
. assert nhwhite+nhblack+hispanic+nhother==1 if ~missing(race, hisplatino) 

. 
. ** Generate education categories
. ** Assign new values to education index variable to correspond with each level of education,
. ** and generate dummies to represent each of these levels
. recode education min/8=1 9=2 10/12=3 13=4 14/max=5, gen(edu_ix) 
(4592 differences between education and edu_ix)

. 
. ** Check to make sure all education level values are now 1-5
. assert    edu_ix <= 5 & edu_ix>=1 if edu_ix<.

. 
. ** Label education index variable and all of the dummies representing education levels
. label var edu_ix "Education Index, 1-5 Scale"

. label def edu_ix 1 "HS Dropout" 2 "High School" 3 "Some College" 4 "Bachelors" 5 "Graduate Degree" 

. label val edu_ix edu_ix

. 
. ** Generate new binary variables representing each education level
. ** and check for any missing values
. gen byte ed_dropout = edu_ix==1 if edu_ix <.
(2 missing values generated)

. gen byte ed_hschool = edu_ix==2 if edu_ix <.
(2 missing values generated)

. gen byte ed_somecol = edu_ix==3 if edu_ix <.
(2 missing values generated)

. gen byte ed_college = edu_ix==4 if edu_ix <.
(2 missing values generated)

. gen byte ed_graduat = edu_ix==5 if edu_ix <.
(2 missing values generated)

. 
. ** Label each of the new education variables
. ** and label them as noyes variables
. label var ed_dropout  "HS Dropout"

. label var ed_hschool  "High School"

. label var ed_somecol  "Some College"

. label var ed_college  "Bachelor's Degree"

. label var ed_graduat  "Graduate Degree"

. 
. label val ed_dropout noyes

. label val ed_hschool noyes

. label val ed_somecol noyes

. label val ed_college noyes

. label val ed_graduat noyes

. 
. ** Generate family income categories
. ** Assign new values (1-5) to hhincome variable to correspond with each increasing 
. ** level of household income. Generate dummies to represent each of these levels
. recode hhincome min/7=1 8/11=2 12/13=3 14=4 15/16=5, gen(income_cat)
(4391 differences between hhincome and income_cat)

. 
. ** Check to make sure all household income level values are now between 1-5
. assert income_cat <= 5 & income_cat>=1 if income_cat<.

. 
. ** Label household income variable and each of the 5 different income levels
. label var income_cat "Income categories (1-5)"

. label def income_cat 1 "HhInc <25k" 2 "HhInc 25-50k" 3 "HhInc 50-75k" 4 "HhInc 75-100k" 5 "HhInc >=100k"

. label val income_cat income_cat

. 
. ** Also use the more detailed HRS-level income variable
. ** Assign new values (1-5) to h12itot variable to correspond with each increasing 
. ** level of household income according to hrs data. Generate dummies for each of 
. ** these levels under name hrsincome_cat
. recode h12itot min/24999 = 1 25000/49999=2 50000/74999=3 75000/99999=4 100000/max=5, gen(hrsincome_cat)
(4388 differences between h12itot and hrsincome_cat)

. 
. ** Check to make sure all hrs household income level values are now 1-5 or missing
. assert hrsincome_cat==1|hrsincome_cat==2|hrsincome_cat==3|hrsincome_cat==4|hrsincome_cat==5|missing(hrsincome_cat)

. 
. ** Use the more detailed HRS income measure if available
. ** Replace income_cat data with all corresponding non-missing hrs data
. replace income_cat = hrsincome_cat if ~missing(hrsincome_cat)
(2,360 real changes made)

. 
. ** Create dummies with more intuitive names for household income categories 
. ** also check for any missing values 
. gen byte hinc_lt25   = income_cat==1 if income_cat <.
(2 missing values generated)

. gen byte hinc_25_50  = income_cat==2 if income_cat <.
(2 missing values generated)

. gen byte hinc_50_75  = income_cat==3 if income_cat <.
(2 missing values generated)

. gen byte hinc_75_100 = income_cat==4 if income_cat <.
(2 missing values generated)

. gen byte hinc_ge100  = income_cat==5 if income_cat <.
(2 missing values generated)

. 
. ** Label new household income dummies
. label var hinc_lt25   "Hh Income: Below 25k"

. label var hinc_25_50  "Hh Income: 25k-50k"

. label var hinc_50_75  "Hh Income: 50k-75k"

. label var hinc_75_100 "Hh Income: 75k-100k"

. label var hinc_ge100  "Hh Income: 100k or more"

. 
. ** Generate and label household size variable & indicator for kids in the hh variable
. gen byte hhsize=1

. gen byte hhkids=0

. label var hhsize "HH size counted from roster"

. label var hhkids "# of kids in HH"

. 
. ** Loop through all 10 (max) potential household members using foreach loop
. ** if that household member exists, add 1 to hhsize variable
. ** if they exist and are below 18 y.o., add 1 to hhkids
. foreach num of numlist 1/10 {
  2.         replace hhsize = hhsize + (hhmemberin_`num'==1)
  3.         replace hhkids = hhkids + (hhmemberin_`num'==1)*(hhmemberage_`num'<18)
  4. }
(3,608 real changes made)
(376 real changes made)
(1,874 real changes made)
(1,103 real changes made)
(1,108 real changes made)
(768 real changes made)
(493 real changes made)
(367 real changes made)
(183 real changes made)
(124 real changes made)
(56 real changes made)
(36 real changes made)
(23 real changes made)
(15 real changes made)
(10 real changes made)
(5 real changes made)
(2 real changes made)
(1 real change made)
(1 real change made)
(0 real changes made)

. ** Make sure hhsiz variable equals the reported household size plus the respondent (+1)
. assert hhsize==hhmembernumber+1  if hhmembernumber <.

. 
. ** Generate household size dummies for sizes of 1, 2, 3, and 4 or more persons
. ** including the respondent in this count
. gen byte hhsiz_1  = hhsize==1 if hhsize <. 

. gen byte hhsiz_2  = hhsize==2 if hhsize <.  

. gen byte hhsiz_3  = hhsize==3 if hhsize <.  

. gen byte hhsiz_4p = hhsize>=4 if hhsize <. 

. 
. ** Make sure that all respondents' households are counted exactly once
. assert hhsiz_1 + hhsiz_2 + hhsiz_3 + hhsiz_4p == 1

. 
. ** Label dummies for household size including respondent 
. label var hhsiz_1  "HhSize=1"

. label var hhsiz_2  "HhSize=2"

. label var hhsiz_3  "HhSize=3"

. label var hhsiz_4p "HhSize>=4"

. 
. ** Generate and label binary variable for any kids in the household and check for missing values
. ** 1 means there is at least 1 kid in house, 0 means no kids
. gen anykids = hhkids >=1 if hhkids <.

. label var anykids "Any kids present in HH"

. label val anykids noyes

. 
. 
. 
. 
. 
.                         
. //  ***************************************************************************************
. //                      RECODE AND RELABEL IN FINANCIAL LITERACY AND COGNITION DATA
. //  ***************************************************************************************
. 
. ** Code the standard Lusardi & Mitchell financial literacy questions
. ** -----------------------------------------------------------------
. 
. **  Suppose you had $100 in a savings account and the interest rate was 2% per year. After 5 years, how much will your money grow?
. gen l001_correct = l001==1 if l001<. // More than $102 is correct answer
(293 missing values generated)

. label var l001_correct "FinLit compounding 2% correct"

. 
. **  Suppose you had $100 in a savings account and the interest rate was 20% per year and you never withdrew money. 
. **  After five years, how much would you have total?
. gen l002_correct = l002==1 if l002<. 
(297 missing values generated)

. label var l002_correct "FinLit compounding 20% correct"

. 
. ** Imagine interest rate was 1% and inflation was 2%. After 1 year how much could you buy?
. gen l003_correct = l003==3 if l003<. 
(300 missing values generated)

. label var l003_correct "FinLit low inflation correct"

. 
. ** Assume a friend inherits $10,000 today and sibling inherits $10,000 in 3 years from now. Who is richer today?
. gen l004_correct = l004==1 if l004<.
(296 missing values generated)

. label var l004_correct "FinLit discounting correct"

. 
. ** Suppose that in the year 2020, your income has doubled and prices have doubled. How much can you buy?
. gen l005_correct = l005==2 if l005<. 
(297 missing values generated)

. label var l005_correct "FinLit high inflation correct"

. 
. ** Which defines the functions of the stock market?
. gen d001_correct = d001==3 if d001<. 
(236 missing values generated)

. label var d001_correct "FinLit stock market description correct"

.                         
. ** Describe a mutual fund
. gen d002_correct = d002==2 if d002<. 
(238 missing values generated)

. label var d002_correct "FinLit mutual fund description correct"

. 
. ** If interest rates rise/fall, what happens to bond prices? (Rise->bond prices fall; Fall->bond prices rise
. gen p001_correct = ((p001_randomizer==1 & p001==2)|(p001_randomizer==2 & p001==1)) if p001<. 
(235 missing values generated)

. label var p001_correct "FinLit bond price correct"

. 
. ** Safety of purchasing single company or stock market fund? Buying single company/stock mutual provides safer return than single company/
> stock?
. gen p002_correct = ((p002_randomizer==1 & p002==2)|(p002_randomizer==2 & p002==1)) if p002<. 
(238 missing values generated)

. label var p002_correct "FinLit diversification correct"

. 
. ** What is riskier, stocks or bonds? 
. gen p003_correct = ((p003_randomizer==1 & p003==1)|(p003_randomizer==2 & p003==2)) if p003<. 
(237 missing values generated)

. label var p003_correct "FinLit risk bonds vs stocks correct"

. 
. ** Considering a long period, what normally gives the highest return?
. gen p004_correct = p004==3 if p004<. 
(235 missing values generated)

. label var p004_correct "FinLit risk long-run returns correct"

. 
. ** Normally, which asset below displays the highest fluctuations over time?
. gen p005_correct = p005==3 if p005<. 
(236 missing values generated)

. label var p005_correct "FinLit risk asset fluctuations correct"

. 
. ** When an investor spreads his money..does the risk..?
. gen p006_correct = p006==2 if p006<. 
(236 missing values generated)

. label var p006_correct "FinLit spreading money correct"

. 
. ** Housing prices in the US can never go down?
. gen p007_correct = p007==2 if p007<. 
(235 missing values generated)

. label var p007_correct "FinLit housing prices correct"

. 
. ** drop the underlying variables (not used anymore)
. drop l001 l002 l003 l004 l005 d001 d002 p001_randomizer p001 p002 p002_randomizer p003 p003_randomizer p004 p005 p006 p007

. 
. ** Generate variable to represent total # of correct FinLit answers
. gen total_fin_lit = l001_correct + ///
>                     l002_correct + ///
>                     l003_correct + ///
>                     l004_correct + ///
>                     l005_correct + ///
>                     d001_correct + ///
>                     d002_correct + ///
>                     p001_correct + ///
>                     p002_correct + ///
>                     p003_correct + ///
>                     p004_correct + ///
>                     p005_correct + ///
>                     p006_correct + ///
>                     p007_correct
(306 missing values generated)

.              
. label var total_fin_lit "FinLit questions correct 0-14"

.  
. ** Calculate percentage of correct responses for each person
. gen fin_lit_percent=total_fin_lit/14
(306 missing values generated)

. label var fin_lit_percent "Fraction FinLit questions correct 0-1" 

. 
. 
. ** Code the numeracy and literacy cognition measures
. ** -------------------------------------------------
. 
. ** relabel the IRT-based numeracy score (from uas1)
. rename uas1cog cog_numbers1

. label var cog_numbers1 "Cog: Numeracy Score"

. 
. 
. ** Rename number series test data (from uas42)
. rename uas42cog cog_numbers2

. label var cog_numbers2 "Cog: Number Series Score"

. 
. 
. ** Rename picture vocabulary test data (from uas43)
. rename uas43cog cog_verbal1

. label var cog_verbal1  "Cog: Picture Vocabulary Score"

. 
. ** Rename verbal analogies test data (from uas44)
. rename uas44cog cog_verbal2

. label var cog_verbal2  "Cog: Verbal Analogies Score"

. 
.         
.         
.         
.         
.         
. 
. // *******************************************************************************************
. // CREATE COGNITION INDEX (for ease of interpretation in regressions)
. // *******************************************************************************************
.                         
. ** Generate indicator for all cognition measures being nonmissing 
. gen cogn_nonmiss = ~missing(total_fin_lit, cog_numbers1, cog_numbers2, cog_verbal1, cog_verbal2)

. 
. ** Standardize the cognition subscores so that they are comparable
. egen cog_fin = std(total_fin_lit)     if cogn_nonmiss
(490 missing values generated)

. egen cog_n1  = std(cog_numbers1)      if cogn_nonmiss
(490 missing values generated)

. egen cog_n2  = std(cog_numbers2)      if cogn_nonmiss
(490 missing values generated)

. egen cog_v1  = std(cog_verbal1)       if cogn_nonmiss
(490 missing values generated)

. egen cog_v2  = std(cog_verbal2)       if cogn_nonmiss
(490 missing values generated)

. 
. ** Label standardized cognition subscore variables
. label var cog_fin "Financial literacy (standardized)"

. label var cog_n1 "Cog: Numeracy (standardized)"

. label var cog_n2 "Cog: Number series (standardized)"

. label var cog_v1 "Cog: Picture vocabulary (standardized)"

. label var cog_v2 "Cog: Verbal analogies (standardized)"

. 
. ** Create the cognition index as the first principal component of the subscores
. ** and standardize it
. pca cog_fin cog_n1 cog_n2 cog_v1 cog_v2   if cogn_nonmiss

Principal components/correlation                 Number of obs    =      4,106
                                                 Number of comp.  =          5
                                                 Trace            =          5
    Rotation: (unrotated = principal)            Rho              =     1.0000

    --------------------------------------------------------------------------
       Component |   Eigenvalue   Difference         Proportion   Cumulative
    -------------+------------------------------------------------------------
           Comp1 |      2.91851       2.2158             0.5837       0.5837
           Comp2 |      .702709     .0854842             0.1405       0.7242
           Comp3 |      .617225      .197959             0.1234       0.8477
           Comp4 |      .419266     .0769794             0.0839       0.9315
           Comp5 |      .342287            .             0.0685       1.0000
    --------------------------------------------------------------------------

Principal components (eigenvectors) 

    ------------------------------------------------------------------------------
        Variable |    Comp1     Comp2     Comp3     Comp4     Comp5 | Unexplained 
    -------------+--------------------------------------------------+-------------
         cog_fin |   0.4512    0.1897   -0.6033   -0.5112    0.3676 |           0 
          cog_n1 |   0.4856   -0.3165   -0.2260    0.0149   -0.7827 |           0 
          cog_n2 |   0.4618   -0.4828    0.0153    0.5614    0.4881 |           0 
          cog_v1 |   0.4005    0.7919    0.1210    0.4339   -0.0984 |           0 
          cog_v2 |   0.4323   -0.0603    0.7551   -0.4848    0.0654 |           0 
    ------------------------------------------------------------------------------

. predict tmp                               if cogn_nonmiss
(score assumed)
(4 components skipped)

Scoring coefficients 
    sum of squares(column-loading) = 1

    ----------------------------------------------------------------
        Variable |    Comp1     Comp2     Comp3     Comp4     Comp5 
    -------------+--------------------------------------------------
         cog_fin |   0.4512    0.1897   -0.6033   -0.5112    0.3676 
          cog_n1 |   0.4856   -0.3165   -0.2260    0.0149   -0.7827 
          cog_n2 |   0.4618   -0.4828    0.0153    0.5614    0.4881 
          cog_v1 |   0.4005    0.7919    0.1210    0.4339   -0.0984 
          cog_v2 |   0.4323   -0.0603    0.7551   -0.4848    0.0654 
    ----------------------------------------------------------------

. egen cognix_pca=std(tmp)
(490 missing values generated)

. drop tmp

. label var cognix_pca "Principal Comp. Cognition Score (standardized)"

. sum  cognix_pca                           if cogn_nonmiss, d

       Principal Comp. Cognition Score (standardized)
-------------------------------------------------------------
      Percentiles      Smallest
 1%    -2.329812      -3.804923
 5%    -1.730106      -3.711271
10%    -1.347875       -3.63989       Obs               4,106
25%    -.7199962      -3.466566       Sum of Wgt.       4,106

50%     .0583676                      Mean           4.55e-10
                        Largest       Std. Dev.             1
75%     .7709345       2.089298
90%     1.282866        2.09919       Variance              1
95%      1.51985       2.172335       Skewness      -.2943381
99%     1.847662       2.185315       Kurtosis       2.583666

.                         
. ** Create quartiles of cognition by which to split the sample
. xtile cognix_xtile4 = cognix if ~missing(cognix), nq(4)

. tab cognix_xtile4, m

4 quantiles |
 of cognix  |      Freq.     Percent        Cum.
------------+-----------------------------------
          1 |      1,027       22.35       22.35
          2 |      1,026       22.32       44.67
          3 |      1,027       22.35       67.01
          4 |      1,026       22.32       89.34
          . |        490       10.66      100.00
------------+-----------------------------------
      Total |      4,596      100.00

. label var cognix_xtile4 "Quartiles of the PCA cognition index"

. 
.                         
. **  As a check: simple average (or sum) of all standardized scores from tests (finlit and cognition), standardized 
. gen tmp = cog_fin+cog_n1+cog_n2+cog_v1+cog_v2 if cogn_nonmiss
(490 missing values generated)

. egen cognix_avg = std(tmp)
(490 missing values generated)

. drop tmp

. label var cognix_avg "Average Cognition Score (standardized)"

. sum  cognix_avg                          if cogn_nonmiss, d

           Average Cognition Score (standardized)
-------------------------------------------------------------
      Percentiles      Smallest
 1%     -2.34872      -3.879393
 5%    -1.740253      -3.770149
10%    -1.339848      -3.644733       Obs               4,106
25%    -.7190309      -3.548534       Sum of Wgt.       4,106

50%     .0676271                      Mean          -4.55e-10
                        Largest       Std. Dev.             1
75%     .7730019       2.083461
90%     1.281757       2.095117       Variance              1
95%     1.510064       2.180609       Skewness      -.3137358
99%     1.844046       2.189435       Kurtosis       2.609102

.                         
.                         
.         
.         
.                         
.                         
.                         
. // *******************************************************************************************
. // CREATE EXTRA DATA VARIABLES JUST FOR APPENDIX TABLES 8-10
. // *******************************************************************************************
. ** The data below is used only for specification 7 of Appendix 
. ** Table 8, Appendix Table 9, and Appendix Table 10
. ** All these variables relate to cognition or familiarity with financial instruments
. 
. 
. ** Score the Adult Decision Making Competence (ADMC) measures on 3 dimensions, following 
. ** Sinayev and Peters, 2015.  Then we do again following Bruine(2007)
. ** -------------------------------------------------------------------------------------
. 
. ** Scoring of Consistency in Risk Perception 
. ** Consistency in Risk Perception assesses the ability to follow probability rules. 
. gen time_conjunct=0 if !missing(admc1)
(13 missing values generated)

. foreach x of numlist 13 14 16 17 19/22 { //taking the "in five years" var and comparing to the relevant "next year" var
  2. local y=`x'-10 //for each five year variable, the associated "next year" is 10 under, 3 and 13, 4 and 14, 6 and 16...
  3. replace time_conjunct=time_conjunct+1 if admc`y'>admc`x' //add 1 to score if greater probability assigned to next year event
  4. }
(616 real changes made)
(541 real changes made)
(1,065 real changes made)
(346 real changes made)
(446 real changes made)
(2,223 real changes made)
(522 real changes made)
(2,558 real changes made)

. 
. ** Scoring of Framing Inconsistency
. ** complementary events (should add to 100% probability) are added, to see if they're within 10 points of 100
. gen framing_inconsistency=0 if !missing(admc1)
(13 missing values generated)

. gen framing_inconsistency_bruine=0 if !missing(admc1)
(13 missing values generated)

. gen add1=admc3+admc12 //3 and 12 are complementary events, so here we're adding them together
(12 missing values generated)

. gen add2=admc7 + admc10 //7 and 10 are complementary events, so here we're adding them together
(13 missing values generated)

. gen add3=admc13 + admc22 //13 and 22 are complementary events, so here we're adding them together
(15 missing values generated)

. gen add4=admc17 + admc20 //17 and 20 are complementary events, so here we're adding them together
(14 missing values generated)

. 
. foreach x of numlist 1/4 {
  2. replace framing_inconsistency=framing_inconsistency+1 if add`x'<90 | add`x'>110 //add 1 to score if not within window around 100
  3. replace framing_inconsistency_bruine=framing_inconsistency_bruine+1 if add`x'<100 | add`x'>100 //add 1 to score if not 100 (bruine scor
> ing procedure has no window)
  4. drop add`x'
  5. }
(1,526 real changes made)
(3,084 real changes made)
(778 real changes made)
(1,831 real changes made)
(1,869 real changes made)
(3,100 real changes made)
(1,160 real changes made)
(2,229 real changes made)

. 
. ** Scoring of Subset Fallacy
. ** add 1 to score if a subset event is marked as higher probability than a superset event
. gen subset_fallacy=0 if !missing(admc1)
(13 missing values generated)

. replace subset_fallacy=subset_fallacy+1 if admc4>admc11 //comparing subset to superset event
(277 real changes made)

. replace subset_fallacy=subset_fallacy+1 if admc14>admc21 //comparing subset to superset event
(337 real changes made)

. 
. ** Re-scoring of Consistency in Risk Perception -- Bruine(2007). 
. ** This is percent correct out of all 14 probability pairs
. ** We also use the much narrower scoring for the framing_inconsistency component
. ** To clarify, admc_bruine is the alternative to using each of the above 3 sub-scores
. gen admc_bruine  = 1 - (framing_inconsistency_bruine + time_conjunct + subset_fallacy)/14
(13 missing values generated)

. 
. ** also create the subscores on a 0-1 scale, and a combined score on 0-1 scale
. gen admc_time    = 1 - time_conjunct/8
(13 missing values generated)

. gen admc_frame   = 1 - framing_inconsistency/4
(13 missing values generated)

. gen admc_subset  = 1 - subset_fallacy/2
(13 missing values generated)

. gen admc_totscore= 1 - (framing_inconsistency + time_conjunct + subset_fallacy)/14
(13 missing values generated)

. 
. ** and standardize them for comparability
. foreach x in bruine time frame subset totscore {
  2.   egen     admc_S`x' = std(admc_`x')
  3. }
(13 missing values generated)
(13 missing values generated)
(13 missing values generated)
(13 missing values generated)
(13 missing values generated)

. 
. ** Show scores
. sum admc_*

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
 admc_bruine |      4,583    .7011471    .1592232   .0714286          1
   admc_time |      4,583    .7731562    .1554661          0          1
  admc_frame |      4,583    .7090879    .2749763          0          1
 admc_subset |      4,583    .9330133    .1998117          0          1
admc_totsc~e |      4,583    .7776877    .1394857   .1428571          1
-------------+---------------------------------------------------------
admc_Sbruine |      4,583    8.18e-09           1  -3.954943   1.876944
  admc_Stime |      4,583   -3.25e-09           1  -4.973148    1.45912
 admc_Sframe |      4,583   -1.30e-08           1  -2.578724   1.057953
admc_Ssubset |      4,583   -7.60e-09           1  -4.669462    .335249
admc_Stots~e |      4,583   -8.24e-09           1  -4.551223     1.5938

. 
. ** Dummy out missing values
. foreach x in bruine time frame subset totscore Sbruine Stime Sframe Ssubset Stotscore {
  2.   gen     admc__`x'_m = missing(admc_`x')
  3.   gen     admc__`x'   = admc_`x'
  4.   replace admc__`x'   = 0  if admc__`x'_m
  5. }
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)
(13 missing values generated)
(13 real changes made)

. 
. ** These variables are not used anymore
. drop admc1-admc22

. 
. 
. 
. ** Now we create index of SS literacy questions (Social Security Attitutes)
. ** The questions are q12_ and q10*_ (several sub-questions of q10 exist)
. ** Q12 is multiple choice and q10* are all True/False
. ** The data already includes Q**_correct for each question as a 0/1 variable, which is what we use below 
. ** -----------------------------------------------------------------------------------------------------
. egen miss=rowmiss(q12_correct q10*_correct) // mark people with missing data

. egen ss_literacy=rowmean(q12_correct q10*_correct) if miss==0  //creat a prop. correct, only for everyone who has all measures
(304 missing values generated)

. drop miss

. 
. ** Show scores
. d   q2c s7a q6a q12* q10* ss_literacy

              storage   display    value
variable name   type    format     label      variable label
--------------------------------------------------------------------------------------------------------------------------------------------
q2c             byte    %26.0f     uas16_vl172
                                              How the Social Security system works
s7a             byte    %5.0f      uas16_vl119
                                              currently receive Social Security benefits
q6a             byte    %25.0f     uas16_vl183
                                              How confident are you that Social Security retirement benefits will be there for
q12             byte    %30.0f     uas16_vl203
                                              automatically deducted
q12_correct     byte    %9.0g                 
q10a            byte    %7.0f      uas16_vl198
                                              benefits if their spouse qualifies for Soci
q10b            byte    %7.0f      uas16_vl199
                                              Social Security benefits are not affected by the age at which someone starts cla
q10c            byte    %7.0f      uas16_vl200
                                              Social Security benefits are adjusted for inflation.
q10d            byte    %7.0f      uas16_vl201
                                              Social Security benefits have to be claimed as soon as someone retires.
q10e            byte    %7.0f      uas16_vl202
                                              Retired people who continue to earn income from working or investments may have
q10f            byte    %7.0f      uas16_vl237
                                              Social Security is paid for by a tax placed on both workers and employers.
q10g            byte    %7.0f      uas16_vl238
                                              Workers who pay Social Security taxes are entitled to Social Security disability
q10h            byte    %7.0f      uas16_vl239
                                              If a worker who pays Social Security taxes dies, any of his/her children under a
q10i            byte    %7.0f      uas16_vl240
                                              If a worker who pays Social Security taxes dies, his/her spouse may claim Social
q10a_correct    byte    %9.0g                 
q10b_correct    byte    %9.0g                 
q10c_correct    byte    %9.0g                 
q10d_correct    byte    %9.0g                 
q10e_correct    byte    %9.0g                 
q10f_correct    byte    %9.0g                 
q10g_correct    byte    %9.0g                 
q10h_correct    byte    %9.0g                 
q10i_correct    byte    %9.0g                 
ss_literacy     float   %9.0g                 

. sum q2c s7a q6a q12* q10* ss_literacy

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
         q2c |      4,281    2.309974    .8368924          1          4
         s7a |      4,336    1.756227    .4294069          1          2
         q6a |      3,273     3.04308    1.015606          1          5
         q12 |      4,331    2.287462    1.420983          1          4
 q12_correct |      4,331     .533595    .4989277          0          1
-------------+---------------------------------------------------------
        q10a |      4,307    1.171813    .3772619          1          2
        q10b |      4,305      1.8741    .3317753          1          2
        q10c |      4,306    1.356712    .4790843          1          2
        q10d |      4,303    1.817802    .3860529          1          2
        q10e |      4,305    1.245064     .430175          1          2
-------------+---------------------------------------------------------
        q10f |      4,301    1.146245    .3533928          1          2
        q10g |      4,306    1.092429    .2896645          1          2
        q10h |      4,306    1.138179    .3451281          1          2
        q10i |      4,303    1.645364    .4784586          1          2
q10a_correct |      4,307    .8281867    .3772619          0          1
-------------+---------------------------------------------------------
q10b_correct |      4,305    .8740999    .3317753          0          1
q10c_correct |      4,306    .6432884    .4790843          0          1
q10d_correct |      4,303    .8178015    .3860529          0          1
q10e_correct |      4,305    .7549361     .430175          0          1
q10f_correct |      4,301    .8537549    .3533928          0          1
-------------+---------------------------------------------------------
q10g_correct |      4,306    .9075708    .2896645          0          1
q10h_correct |      4,306    .8618207    .3451281          0          1
q10i_correct |      4,303    .6453637    .4784586          0          1
 ss_literacy |      4,292    .7724371    .1685679         .1          1

. 
. tab q2c 

   How the Social Security |
              system works |      Freq.     Percent        Cum.
---------------------------+-----------------------------------
      1 Very knowledgeable |        632       14.76       14.76
  2 Somewhat knowledgeable |      2,099       49.03       63.79
   3 Not too knowledgeable |      1,141       26.65       90.45
4 Not at all knowledgeable |        409        9.55      100.00
---------------------------+-----------------------------------
                     Total |      4,281      100.00

. tab s7a 

  currently |
    receive |
     Social |
   Security |
   benefits |      Freq.     Percent        Cum.
------------+-----------------------------------
      1 Yes |      1,057       24.38       24.38
       2 No |      3,279       75.62      100.00
------------+-----------------------------------
      Total |      4,336      100.00

. tab q6a 

    How confident are you |
     that Social Security |
 retirement benefits will |
             be there for |      Freq.     Percent        Cum.
--------------------------+-----------------------------------
         1 Very confident |        229        7.00        7.00
     2 Somewhat confident |        791       24.17       31.16
3 Only a little confident |      1,005       30.71       61.87
   4 Not at all confident |      1,106       33.79       95.66
              5 Dont know |        142        4.34      100.00
--------------------------+-----------------------------------
                    Total |      3,273      100.00

. tab q12

        automatically deducted |      Freq.     Percent        Cum.
-------------------------------+-----------------------------------
     1 Medicare Part B premium |      2,311       53.36       53.36
2 Premium for Medigap policies |         81        1.87       55.23
                3 Income taxes |        322        7.43       62.66
                   4 Dont know |      1,617       37.34      100.00
-------------------------------+-----------------------------------
                         Total |      4,331      100.00

. 
. ** prefix these questions with ass (Attitudes on Social Security)
. gen ssa_literacy  = ss_literacy
(304 missing values generated)

. gen ssa_knowledge = q2c
(315 missing values generated)

. gen ssa_confident = q6a
(1,323 missing values generated)

. 
. ** code more knowledge as higher value (on 0-1 scale)
. recode ssa_knowledge 5=. 4=0 3=0.33 2=0.67 1=1
(ssa_knowledge: 3649 changes made)

. 
. ** code more confident as higher value (on 0-1 scale), and "don't know" as missing
. recode ssa_confident 5=. 4=0 3=0.33 2=0.67 1=1
(ssa_confident: 3044 changes made)

. 
. ** and standardize them for comparability
. foreach x in literacy knowledge confident {
  2.   egen     ssa_S`x' = std(ssa_`x')
  3. }
(304 missing values generated)
(315 missing values generated)
(1465 missing values generated)

. 
. ** Dummy out missing values
. foreach x in literacy knowledge confident Sliteracy Sknowledge Sconfident {
  2.   gen     ssa__`x'_m = missing(ssa_`x')
  3.   gen     ssa__`x'   = ssa_`x'
  4.   replace ssa__`x'   = 0  if ssa__`x'_m
  5. }
(304 missing values generated)
(304 real changes made)
(315 missing values generated)
(315 real changes made)
(1,465 missing values generated)
(1,465 real changes made)
(304 missing values generated)
(304 real changes made)
(315 missing values generated)
(315 real changes made)
(1,465 missing values generated)
(1,465 real changes made)

. 
. sum ssa_*

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
ssa_literacy |      4,292    .7724371    .1685679         .1          1
ssa_knowle~e |      4,281    .5640878    .2803131          0          1
ssa_confid~t |      3,131    .3483296    .3165454          0          1
ssa_Sliter~y |      4,292   -5.59e-09           1  -3.989118   1.349978
ssa_Sknowl~e |      4,281    3.69e-09           1  -2.012349    1.55509
-------------+---------------------------------------------------------
ssa_Sconfi~t |      3,131    6.25e-09           1   -1.10041   2.058695
ssa__liter~m |      4,596    .0661445    .2485615          0          1
ssa__liter~y |      4,596    .7213446    .2517906          0          1
ssa__knowl~m |      4,596    .0685379    .2526941          0          1
ssa__knowl~e |      4,596    .5254265    .3057891          0          1
-------------+---------------------------------------------------------
ssa__confi~m |      4,596    .3187554    .4660447          0          1
ssa__confi~t |      4,596    .2372977    .3075838          0          1
ssa__Slite~m |      4,596    .0661445    .2485615          0          1
ssa__Slite~y |      4,596   -5.22e-09    .9663545  -3.989118   1.349978
ssa__Sknow~m |      4,596    .0685379    .2526941          0          1
-------------+---------------------------------------------------------
ssa__Sknow~e |      4,596    3.44e-09    .9651151  -2.012349    1.55509
ssa__Sconf~m |      4,596    .3187554    .4660447          0          1
ssa__Sconf~t |      4,596    4.26e-09    .8253334   -1.10041   2.058695

. 
. ** no longer needed
. drop q2c s7a q6a q12* q10* ss_literacy

. 
. 
. 
. ** Create an index of ability and comfort with financial planning
. ** ---------------------------------------------------------------------------------
. ** ch009a "Have enough information"             --> agree (1) = more ability/comfort  --> REVERSE code
. ** ch009b "Not interested in learning about ret planning". Unclear what this means. It could be a very able person
. **         who doesn't need more info, or someone with their head in the sand         --> OMIT
. ** ch009c "don't know best source of info"      --> agree (1) = less ability/comfort  --> Regular code
. ** ch009d "comfortable with online bank"        --> agree (1) = more ability/comfor   --> REVERSE code
. ** ch009e "comfortable with online search ret"  --> agree (1) = more ability/comfort  --> REVERSE code
. ** ch009f "comfortable with online search gov"  --> agree (1) = more ability/comfort  --> REVERSE code
. 
. ** standarize the index
. egen aplan_index =std(-ch009a+ch009c-ch009d-ch009e-ch009f)
(1100 missing values generated)

. 
. sum aplan_index, d

                   Standardized values of
         (-ch009a+ch009c-ch009d-ch009e-ch009f)     
-------------------------------------------------------------
      Percentiles      Smallest
 1%    -2.457073       -2.78208
 5%     -1.80706       -2.78208
10%    -1.482053       -2.78208       Obs               3,496
25%    -.5070328       -2.78208       Sum of Wgt.       3,496

50%     .1429807                      Mean           3.47e-09
                        Largest       Std. Dev.             1
75%     .7929941       2.093021
90%     1.118001       2.093021       Variance              1
95%     1.443008       2.093021       Skewness       -.265877
99%     2.093021       2.093021       Kurtosis       2.884061

. 
. ** Dummy out missing values
. foreach x in index {
  2.   gen     aplan__`x'_m = missing(aplan_`x')
  3.   gen     aplan__`x'   = aplan_`x'
  4.   replace aplan__`x'   = 0  if aplan__`x'_m
  5. }
(1,100 missing values generated)
(1,100 real changes made)

. 
. sum aplan*

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
 aplan_index |      3,496    3.47e-09           1   -2.78208   2.093021
aplan__ind~m |      4,596    .2393386    .4267262          0          1
aplan__index |      4,596    2.64e-09    .8721292   -2.78208   2.093021

. 
. ** no longer needed
. drop ch009* 

. 
. ** Create indicators for annuity holdings or IRA/KEOGH holdings
. ** ---------------------------------------------------------------------------------
. 
. ** Create indicator variable for annuity holdings (self or spouse)
. tab q273_, m

    R OR SP |
INCOME FROM |
  ANNUITIES |      Freq.     Percent        Cum.
------------+-----------------------------------
      1 Yes |         95        2.07        2.07
       5 No |      4,358       94.82       96.89
          . |        129        2.81       99.70
         .a |         13        0.28       99.98
         .e |          1        0.02      100.00
------------+-----------------------------------
      Total |      4,596      100.00

. gen afin_annuity = q273_==1  if !missing(q273_)
(143 missing values generated)

. 
. ** Create indicator variable for owning IRA or Keogh
. tab q162_, m

     IRA OR |
      KEOGH |      Freq.     Percent        Cum.
------------+-----------------------------------
      1 Yes |      1,596       34.73       34.73
       5 No |      2,859       62.21       96.93
          . |        129        2.81       99.74
         .a |         10        0.22       99.96
         .e |          2        0.04      100.00
------------+-----------------------------------
      Total |      4,596      100.00

. gen afin_irakeogh = q162_==1 if !missing(q162_)
(141 missing values generated)

. 
. ** Dummy out missing values
. foreach x in annuity irakeogh {
  2.   gen     afin__`x'_m = missing(afin_`x')
  3.   gen     afin__`x'   = afin_`x'
  4.   replace afin__`x'   = 0  if afin__`x'_m
  5. }
(143 missing values generated)
(143 real changes made)
(141 missing values generated)
(141 real changes made)

. 
. sum afin*

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
afin_annuity |      4,453    .0213339    .1445112          0          1
afin_irake~h |      4,455    .3582492    .4795397          0          1
afin__annu~m |      4,596     .031114    .1736447          0          1
afin__annu~y |      4,596    .0206701     .142293          0          1
afin__irak~m |      4,596    .0306789    .1724649          0          1
-------------+---------------------------------------------------------
afin__irak~h |      4,596    .3472585    .4761506          0          1

. 
. * no longer needed
. drop q273_ q162_ 

. 
. 
. 
. 
. 
. 
. // ***************************************************************************************
. //      SET GLOBAL VARIABLES FOR REGRESSIONS AND OTHER ANALYSES
. // ***************************************************************************************
. 
. 
. ** Demographics global for balance tests
. global demographics_balance age agesq female married                                   ///
>                                 nhwhite     nhblack nhother hispanic                       ///
>                                 ed_dropout ed_hschool ed_somecol ed_college ed_graduat     ///
>                                 hinc_lt25   hinc_25_50 hinc_50_75 hinc_75_100 hinc_ge100   ///
>                                 hhsiz_1     hhsiz_2 hhsiz_3 hhsiz_4p anykids

.                                                         
.          
. ** Demographic global for regressions
. global demographics         age agesq female married                                   ///
>                                             nhblack nhother hispanic                       ///
>                                 ed_dropout ed_hschool            ed_college ed_graduat     ///
>                                             hinc_25_50 hinc_50_75 hinc_75_100 hinc_ge100   ///
>                                             hhsiz_2 hhsiz_3 hhsiz_4p anykids

.                                                                                 
. ** Experimental controls global for balance tests
. global exp_controls_balance sell_first                                                        ///
>                                 ls_startvalue_1 ls_startvalue_2 ls_startvalue_3 ls_first          ///
>                                 ss_benefit_1    ss_benefit_2 ss_benefit_3 ss_benefit_4            ///
>                                 vignette_name_1 vignette_name_2 vignette_name_3 vignette_name_4 

. 
. ** Experimental controls for regressions        
. global exp_controls         sell_first                                                        ///
>                                                 ls_startvalue_2 ls_startvalue_3 ls_first          ///
>                                                 ss_benefit_2 ss_benefit_3 ss_benefit_4            ///
>                                                 vignette_name_2 vignette_name_3 vignette_name_4 

. 
. ** Experimental controls for regressions, but with linear SS benefit variable (rather than 4 dummies)   
. global exp_controls_linben  sell_first                                                        ///
>                                             ls_startvalue_2 ls_startvalue_3 ls_first          ///
>                                             ss_benefit100dollar                               ///
>                                             vignette_name_2 vignette_name_3 vignette_name_4 

.         
. 
. 
. 
. 
. 
. 
. 
. // ***************************************************************************************
. // ***************************************************************************************
. // 
. // SAMPLE SELECTION
. //
. // ***************************************************************************************
. // ***************************************************************************************
. 
. 
. ** Examine missing values in outcome data
. ** See overlap between the two types of missing information
. gen miss_buy    = missing(logbuyprice)

. gen miss_sell   = missing(logsellprice)

. gen miss_spread = miss_buy | miss_sell

. tab miss_buy miss_sell, m

           |       miss_sell
  miss_buy |         0          1 |     Total
-----------+----------------------+----------
         0 |     4,552          5 |     4,557 
         1 |        11         28 |        39 
-----------+----------------------+----------
     Total |     4,563         33 |     4,596 


. 
. ** Check whether missings happen disproportionally for any treatment
. tab treat2x3 if miss_spread

Main treatments, 2x3 design |      Freq.     Percent        Cum.
----------------------------+-----------------------------------
                   Baseline |          7       15.91       15.91
Wide Spread; NO conseq. msg |          5       11.36       27.27
Added Info ; NO conseq. msg |         12       27.27       54.55
 Narrow Spread; CONSEQ. msg |          5       11.36       65.91
 Wide Spread  ; CONSEQ. msg |          7       15.91       81.82
 Added Info   ; CONSEQ. msg |          8       18.18      100.00
----------------------------+-----------------------------------
                      Total |         44      100.00

. 
. ** Examine how often demographic data is missing
. foreach var of varlist age gender education race hisplatino marital income_cat hhsize hhkids {
  2.         qui gen miss_`var'= missing(`var')
  3.         di "Number of missings for `var':"
  4.         count if miss_`var'
  5.         di ""
  6.         qui drop miss_`var'
  7.         }
Number of missings for age:
  7

Number of missings for gender:
  3

Number of missings for education:
  2

Number of missings for race:
  17

Number of missings for hisplatino:
  2

Number of missings for maritalstatus:
  3

Number of missings for income_cat:
  2

Number of missings for hhsize:
  0

Number of missings for hhkids:
  0


. gen     miss_anydemographic = missing(age, gender, education, race, hisplatino, marital, income_cat, hhsize, hhkids)

. tab     miss_anydemographic

miss_anydem |
   ographic |      Freq.     Percent        Cum.
------------+-----------------------------------
          0 |      4,571       99.46       99.46
          1 |         25        0.54      100.00
------------+-----------------------------------
      Total |      4,596      100.00

. 
. ** Examine missing variables for missing cognition data
. ** As we saw before, most of these missings come from missing finlit questions
. gen miss_cognix_pca =   missing(cognix_pca)

. tab miss_cognix_pca

miss_cognix |
       _pca |      Freq.     Percent        Cum.
------------+-----------------------------------
          0 |      4,106       89.34       89.34
          1 |        490       10.66      100.00
------------+-----------------------------------
      Total |      4,596      100.00

. 
. 
. ************* INDICATOR FOR BASELINE SAMPLE ***************
. **
. ** Baseline sample requires that outcome variable, demographics, and cognition are all nonmissing
. ** 
. 
. gen byte basesample = ~(miss_spread | miss_anydemographic | miss_cognix_pca)                    

. 
. ************************************************************
.         
. 
. ** double check that in the demographics global there are no missing values
. foreach var of varlist $demographics {
  2.         qui assert `var' < . if basesample
  3.         }

. 
. 
. ** Make sure that cognition measures are standardized on the basesample
. ** (rather than standardized on a broader sample)
. sum cognix_pca if basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
  cognix_pca |      4,060    .0041276    .9983618  -3.711271   2.185315

. replace cognix_pca = (cognix_pca-r(mean))/r(sd)
(4,106 real changes made)

. 
. sum cognix_avg if basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
  cognix_avg |      4,060    .0041391    .9982539  -3.770149   2.189435

. replace cognix_avg = (cognix_avg-r(mean))/r(sd)
(4,106 real changes made)

. 
. 
. 
. ** Indicator for not completing the entire survey
. ** Note: Some of these are still in our base sample because they did answer the key outcome 
. **       variables and attrited on later questions
. gen byte  attrit = missing_end_date

. label var attrit "Did not complete the entire survey (including the insurance questions that we don't use)"
note: label truncated to 80 characters

. 
. ** For reporting in the paper
. ** From the survey administrators: number of invited participants: 5,521
. 
. ** Percent opening the survey
. count if ~missing_start_date
  4,595

. scalar N_start=r(N)

. di "Percent of all invited panel members that opened the survey: " 100*N_start/5521
Percent of all invited panel members that opened the survey: 83.227676

.  
. 
. ** Percent completing the buy and sell questions conditional on opening the survey
. count if ~miss_spread
  4,552

. scalar N_complete=r(N)

. di "Percent of respondents to the invitation that completed the annuity questions: " 100*N_complete/N_start
Percent of respondents to the invitation that completed the annuity questions: 99.0642

. 
. ** Overall completion rate (for our section)
. di "Percent of all invited panel members that completed the annuity questions: " 100*N_complete/5521
Percent of all invited panel members that completed the annuity questions: 82.448832

. 
. ** Percent of respondents we drop due to missing demographics or cognition questions
. count if miss_anydemographic & ~miss_spread
  24

. di "Percent dropped for missing demographics: " 100*r(N)/N_complete
Percent dropped for missing demographics: .52724077

. 
. count if miss_cognix_pca & ~miss_anydemographic & ~miss_spread
  468

. di "Additional percent dropped for missing cognition measure: " 100*r(N)/N_complete
Additional percent dropped for missing cognition measure: 10.281195

. 
. ** don't have variables take up more space than needed
. quietly compress

. 
. 
. 
. 
. 
. 
. 
. 
.         
.         
.         
. // ***************************************************************************************
. // ***************************************************************************************
. // 
. //  PART 2: ANALYSES
. //
. // ***************************************************************************************
. // ***************************************************************************************
. 
.         
.         
.         
.         
.         
.         
.         
. // ***************************************************************************************
. //                      Figure 1: CDF of Sell Price and Buy Price in the Subsample without Anchoring
. //                              (When buy price and sell price are each shown first)
. // ***************************************************************************************
.    
. ** Export data for first line of figure 1: buy price
. outsheet midbuy  if ~sell_first & basesample using rawdata_fig1a_midbuy.xls, replace

. 
. ** Export data for second line of figure 1: sell price
. outsheet midsell if  sell_first & basesample using rawdata_fig1b_midsell.xls, replace

. 
. ** summarize to get the medians and median regressions to get standard errors
. sum  midbuy  if ~sell_first & basesample, d

                    Buy price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          250            250
10%          250            250       Obs               2,051
25%         1250            250       Sum of Wgt.       2,051

50%         4750                      Mean           21059.61
                        Largest       Std. Dev.      90082.72
75%        13750         750000
90%        23750         750000       Variance       8.11e+09
95%        42500         750000       Skewness       7.430982
99%       750000         750000       Kurtosis       58.34003

. qreg midbuy  if ~sell_first & basesample
Iteration  1:  WLS sum of weighted deviations =   25078320

Iteration  1: sum of abs. weighted deviations =   24310250
Iteration  2: sum of abs. weighted deviations =   19833125

Median regression                                   Number of obs =      2,051
  Raw sum of deviations 1.98e+07 (about 4750)
  Min sum of deviations 1.98e+07                    Pseudo R2     =     0.0000

------------------------------------------------------------------------------
      midbuy |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |       4750   180.4748    26.32   0.000     4396.067    5103.933
------------------------------------------------------------------------------

. 
. sum  midsell if  sell_first & basesample, d

                    Sell price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%         1250            250
10%         2750            250       Obs               2,009
25%         7375            250       Sum of Wgt.       2,009

50%        16250                      Mean           57658.54
                        Largest       Std. Dev.      143146.2
75%        31250         750000
90%        95000         750000       Variance       2.05e+10
95%       350000         750000       Skewness       3.856013
99%       750000         750000       Kurtosis       17.24059

. qreg midsell if  sell_first & basesample
Iteration  1:  WLS sum of weighted deviations =   66086509

Iteration  1: sum of abs. weighted deviations =   65428250
Iteration  2: sum of abs. weighted deviations =   50672625

Median regression                                   Number of obs =      2,009
  Raw sum of deviations 5.07e+07 (about 16250)
  Min sum of deviations 5.07e+07                    Pseudo R2     =     0.0000

------------------------------------------------------------------------------
     midsell |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |      16250   543.2946    29.91   0.000     15184.52    17315.48
------------------------------------------------------------------------------

.    
.    
. 
. // ***************************************************************************************
. //                      Figure 2: CDF of Log Sell Price Minus Log Buy Price
. // ***************************************************************************************
.    
. ** Export data for figure 2: log difference of log(sell) - log(buy)
. outsheet logsellbuydiff if basesample using rawdata_fig2_logdiff.xls, replace

. 
. ** Summary statistics for the mean and median lines in the figure  
. sum logsellbuydiff if basesample, d   

                     Log Sell - Log Buy
-------------------------------------------------------------
      Percentiles      Smallest
 1%    -6.173786      -8.006368
 5%    -3.999034      -8.006368
10%     -2.65926      -8.006368       Obs               4,060
25%    -.1957446      -8.006368       Sum of Wgt.       4,060

50%     .6048546                      Mean           1.009391
                        Largest       Std. Dev.      2.899622
75%     2.564949       8.006368
90%     5.010635       8.006368       Variance       8.407808
95%      6.55108       8.006368       Skewness       .0812005
99%     8.006368       8.006368       Kurtosis       3.486497

.    
.          
.    
.    
. // ***************************************************************************************
. //                      APX Figure A1: CDF of Sell Price and Buy Price in the Entire Baseline Sample
. // ***************************************************************************************
.       
. ** Export data for first line of figure A1: buy price
. outsheet midbuy  if  basesample using rawdata_figA1a_midbuy.xls, replace

. 
. ** Export data for second line of figure A1: sell price
. outsheet midsell if  basesample using rawdata_figA1b_midsell.xls, replace

. 
. ** summarize to get the medians and median regressions to get standard errors
. sum  midbuy  if basesample, d

                    Buy price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          250            250
10%          250            250       Obs               4,060
25%         1250            250       Sum of Wgt.       4,060

50%         5875                      Mean           67327.09
                        Largest       Std. Dev.      189302.6
75%        16250         750000
90%        75000         750000       Variance       3.58e+10
95%       750000         750000       Skewness       3.072675
99%       750000         750000       Kurtosis       10.70886

. qreg midbuy  if basesample
Iteration  1:  WLS sum of weighted deviations =  1.604e+08

Iteration  1: sum of abs. weighted deviations =  1.610e+08
note:  alternate solutions exist
Iteration  2: sum of abs. weighted deviations =  1.328e+08

Median regression                                   Number of obs =      4,060
  Raw sum of deviations 1.33e+08 (about 5875)
  Min sum of deviations 1.33e+08                    Pseudo R2     =     0.0000

------------------------------------------------------------------------------
      midbuy |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |       5875   193.2729    30.40   0.000     5496.079    6253.921
------------------------------------------------------------------------------

. sum  midsell if basesample, d

                    Sell price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          500            250
10%         1750            250       Obs               4,060
25%         6625            250       Sum of Wgt.       4,060

50%        16250                      Mean            79280.7
                        Largest       Std. Dev.      184016.4
75%        32500         750000
90%       175000         750000       Variance       3.39e+10
95%       600000         750000       Skewness       2.947456
99%       750000         750000       Kurtosis       10.27378

. qreg midsell if basesample
Iteration  1:  WLS sum of weighted deviations =  1.721e+08

Iteration  1: sum of abs. weighted deviations =  1.717e+08
note:  alternate solutions exist
Iteration  2: sum of abs. weighted deviations =  1.474e+08

Median regression                                   Number of obs =      4,060
  Raw sum of deviations 1.47e+08 (about 16250)
  Min sum of deviations 1.47e+08                    Pseudo R2     =     0.0000

------------------------------------------------------------------------------
     midsell |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |      16250   483.1824    33.63   0.000      15302.7     17197.3
------------------------------------------------------------------------------

. 
.    
.    
.    
.    
. // ***************************************************************************************
. //                      APX Figure A2: CDF of Sell-Buy Spread
. // ***************************************************************************************
.    
. ** Figure A2: CDF of spread
. ** Export data for figure A2: spread = abs log difference of log(sell) - log(buy)
. outsheet logspread if basesample using rawdata_figA2_spread.xls, replace

. 
.  
. ** Summary statistics for the mean and median lines in the figure  
. sum logspread if basesample, d

                         Log Spread
-------------------------------------------------------------
      Percentiles      Smallest
 1%            0              0
 5%            0              0
10%            0              0       Obs               4,060
25%     .4353181              0       Sum of Wgt.       4,060

50%     1.547563                      Mean           2.214562
                        Largest       Std. Dev.      2.126367
75%     3.563716       8.006368
90%     5.393628       8.006368       Variance       4.521437
95%      6.55108       8.006368       Skewness       1.006302
99%     8.006368       8.006368       Kurtosis       3.133193

. 
. 
. ** Wilcoxon matched-pairs signed-rank test with exact statistics
. signrankex midbuy=midsell if basesample

Wilcoxon signed-rank test

        sign |      obs   sum ranks    expected
-------------+---------------------------------
    positive |     1110   2124498.5     3335189
    negative |     2542   4545879.5     3335189
        zero |      408              
-------------+---------------------------------
         all |     4060     8243830     8243830

Ho: midbuy = midsell 
              t = -20.060
    Prob >= |t| =   0.0000

. 
. 
. 
. 
. 
. 
. 
. 
. 
. 
. 
. 
.         
. // ***************************************************************************************
. //                      TABLE 1: Contains Description of Experimental Design, Not Part of STATA File
. // ***************************************************************************************
. 
. 
. 
. 
. 
. 
.         
. // ***************************************************************************************
. //                      TABLE 2: Descriptive Statistics on the Sell Price, Buy Price, and Spread
. // ***************************************************************************************
. 
. 
. ** Show the summary statistics of the log of the outcome variables, now also including the spread
. ** in a table.
. **
. ** In the tables:
. **    Four rows:
. **       1. "Sell price (log)"
. **       2. "Buy price (log)"
. **       3. "Sell-Buy Spread"
. **       4. "N"  (no. of observations)
. **
. **    Four main columns (but use separate columns in xls for mean and standard deviation, so 7 columns in excel)
. **       1. Sell question asked first
. **          Show the mean and sd in two xls columns for main column 1
. **       2. Buy question asked first
. **          Show the mean and sd in two xls columns for main column 2
. **       3. P-value on the difference in means between col 1 and col 2
. **          This p-value is reported under "P>|t|" on sell_first in the regressions below.
. **       4. Entire baseline sample: show the mean and sd
. **
.    
. ** The means and sd for column 1:
. sum logsellprice logbuyprice logspread if  sell_first & basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
logsellprice |      2,009    9.646391    1.532202   5.521461   13.52783
 logbuyprice |      2,009    9.060307     2.42954   5.521461   13.52783
   logspread |      2,009    2.273642     2.03507          0   8.006368

.    
. ** The means and sd for column 2:
. sum logsellprice logbuyprice logspread if ~sell_first & basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
logsellprice |      2,051    9.708964    1.961954   5.521461   13.52783
 logbuyprice |      2,051    8.284936    1.684851   5.521461   13.52783
   logspread |      2,051    2.156691    2.211111          0   8.006368

. 
. ** The means and sd for column 4:
. sum logsellprice logbuyprice logspread if basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
logsellprice |      4,060    9.678001    1.762509   5.521461   13.52783
 logbuyprice |      4,060    8.668611    2.122283   5.521461   13.52783
   logspread |      4,060    2.214562    2.126367          0   8.006368

. 
. ** regressions for the p-values in column 3 for rows 1-3
. ** The p-value is reported under "P>|t|" on sell_first in the regressions below.
. 
. * P-value for Row 1 
. reg logsellprice sell_first if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(1, 4058)        =       1.29
                                                Prob > F          =     0.2569
                                                R-squared         =     0.0003
                                                Root MSE          =     1.7624

------------------------------------------------------------------------------
             |               Robust
logsellprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  sell_first |  -.0625733   .0551846    -1.13   0.257    -.1707654    .0456188
       _cons |   9.708964   .0433219   224.11   0.000      9.62403    9.793899
------------------------------------------------------------------------------

. 
. * P-Value for Row 2
. reg logbuyprice  sell_first if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(1, 4058)        =     139.10
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0334
                                                Root MSE          =     2.0868

------------------------------------------------------------------------------
             |               Robust
 logbuyprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  sell_first |   .7753707   .0657432    11.79   0.000      .646478    .9042635
       _cons |   8.284936   .0372031   222.69   0.000     8.211997    8.357874
------------------------------------------------------------------------------

. 
. * P-Value for Row 3
. reg logspread    sell_first if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(1, 4058)        =       3.08
                                                Prob > F          =     0.0795
                                                R-squared         =     0.0008
                                                Root MSE          =     2.1258

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  sell_first |   .1169511   .0666723     1.75   0.079    -.0137633    .2476655
       _cons |   2.156691   .0488235    44.17   0.000      2.06097    2.252412
------------------------------------------------------------------------------

. 
.   
.   
.   
.   
.    
. 
.         
. // ***************************************************************************************
. //                      TABLE 3: MAIN REGRESSION; Treatment Effects on the Sell-Buy Spread and its Components
. // ***************************************************************************************
. 
. ** Table 3
. **
. ** This table has 3 main columns, with each main column having two entries, namely
. ** for the coefficient estimate and the standard error.
. **    Col 1: "Sell-Buy Spread"
. **    Col 2: "Sell Price (log)"
. **    Col 3: "Buy Price (log)"
. **
. ** In the rows of the table are the explanatory variables:
. **    Row 1: "Complexity treatment"
. **    Row 2: "Consequence message treatment"
. **    Row 3: "Cognition index"
. **    Row 4: "Sell question first"
. **
. **    Row 5: "P-value on lump-sum starting values" 
. **    Row 6: "P-value on lump-sum shown first" 
. **    Row 7: "P-value on SS benefit amounts"
. **    Row 8: "P-value on vignette names"
. **    
. **    Row 9: "Demographic controls"   (enter "Yes" in each column)
. **   
. **    Row 10: R^2
. **    Row 11: N
. **
. 
. 
. ** ------ Column 1 --------
. ** Estimates for rows 1-4, row 6: the p-value on lump-sum shown first (under "P>|t|"), R2 and N are in the output below 
. reg logspread    any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      24.04
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1568
                                                Root MSE          =     1.9603

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1314293   .0648348     2.03   0.043     .0043173    .2585414
    consequence |  -.1405306   .0616294    -2.28   0.023    -.2613583   -.0197029
     cognix_pca |  -.7876547   .0427562   -18.42   0.000    -.8714804   -.7038289
     sell_first |   .1657743   .0616662     2.69   0.007     .0448745    .2866741
ls_startvalue_2 |   .0632191   .0761058     0.83   0.406    -.0859904    .2124286
ls_startvalue_3 |  -.0021141   .0749548    -0.03   0.978    -.1490669    .1448388
       ls_first |   .0294388   .0616149     0.48   0.633    -.0913606    .1502381
   ss_benefit_2 |    .112784   .0870307     1.30   0.195    -.0578444    .2834124
   ss_benefit_3 |   .0573012   .0842277     0.68   0.496    -.1078317     .222434
   ss_benefit_4 |   .1669049   .0868148     1.92   0.055    -.0033001    .3371099
vignette_name_2 |   .1140634      .0856     1.33   0.183    -.0537599    .2818867
vignette_name_3 |   .0882214   .0877756     1.01   0.315    -.0838674    .2603102
vignette_name_4 |  -.0109552   .0850845    -0.13   0.898    -.1777678    .1558574
            age |   .0247686   .0129824     1.91   0.056    -.0006842    .0502213
          agesq |  -.0147106   .0129063    -1.14   0.254     -.040014    .0105928
         female |   .0854357   .0663126     1.29   0.198    -.0445738    .2154451
        married |   .0965551   .0759943     1.27   0.204    -.0524358    .2455459
        nhblack |   .0281964   .1418699     0.20   0.842    -.2499472      .30634
        nhother |   .0481969   .1215336     0.40   0.692    -.1900762      .28647
       hispanic |   .0810356   .1251889     0.65   0.517    -.1644039    .3264752
     ed_dropout |  -.0573967   .1777837    -0.32   0.747     -.405951    .2911577
     ed_hschool |   .0334476   .0930188     0.36   0.719    -.1489207    .2158159
     ed_college |   .0076616   .0838392     0.09   0.927    -.1567096    .1720329
     ed_graduat |   .0763065   .0954393     0.80   0.424    -.1108073    .2634204
     hinc_25_50 |   .1019141   .1167251     0.87   0.383    -.1269317      .33076
     hinc_50_75 |  -.1660551    .115919    -1.43   0.152    -.3933204    .0612103
    hinc_75_100 |  -.0550164   .1299977    -0.42   0.672    -.3098838     .199851
     hinc_ge100 |  -.2568218   .1099252    -2.34   0.020    -.4723359   -.0413077
        hhsiz_2 |  -.0247458   .0954113    -0.26   0.795    -.2118047    .1623132
        hhsiz_3 |   .1468662   .1305622     1.12   0.261    -.1091079    .4028402
       hhsiz_4p |   .1815883   .1452072     1.25   0.211    -.1030981    .4662747
        anykids |  -.1769094   .1055735    -1.68   0.094    -.3838918    .0300729
          _cons |   1.106907   .3419853     3.24   0.001     .4364261    1.777387
---------------------------------------------------------------------------------

. 
. * Row 5: P-Value on Lump-Sum Starting Values
. testparm ls_startvalue_*

 ( 1)  ls_startvalue_2 = 0
 ( 2)  ls_startvalue_3 = 0

       F(  2,  4027) =    0.47
            Prob > F =    0.6232

. 
. * Row 7: P-Value on SS Benefit Amounts
. testparm ss_benefit_*

 ( 1)  ss_benefit_2 = 0
 ( 2)  ss_benefit_3 = 0
 ( 3)  ss_benefit_4 = 0

       F(  3,  4027) =    1.37
            Prob > F =    0.2489

. 
. * Row 8: P-Value on Vignette Names
. testparm vignette_name_*

 ( 1)  vignette_name_2 = 0
 ( 2)  vignette_name_3 = 0
 ( 3)  vignette_name_4 = 0

       F(  3,  4027) =    1.04
            Prob > F =    0.3748

. 
. 
. 
. ** ------ Column 2 --------
. ** Estimates for rows 1-4, row 6: the p-value on lump-sum shown first (under "P>|t|"), R2 and N are in the output below 
. reg logsellprice any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =       4.60
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0350
                                                Root MSE          =     1.7382

---------------------------------------------------------------------------------
                |               Robust
   logsellprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |    .049683   .0574995     0.86   0.388    -.0630479    .1624138
    consequence |    .010656   .0547079     0.19   0.846    -.0966018    .1179138
     cognix_pca |  -.1883284   .0381627    -4.93   0.000    -.2631484   -.1135084
     sell_first |  -.0425004   .0545079    -0.78   0.436     -.149366    .0643652
ls_startvalue_2 |   .2393336   .0665975     3.59   0.000     .1087658    .3699015
ls_startvalue_3 |   .4842904   .0675399     7.17   0.000     .3518748    .6167059
       ls_first |    -.04367    .054771    -0.80   0.425    -.1510515    .0637114
   ss_benefit_2 |   .0100883   .0754805     0.13   0.894    -.1378952    .1580718
   ss_benefit_3 |  -.0060516   .0739202    -0.08   0.935    -.1509761     .138873
   ss_benefit_4 |  -.1184736   .0798293    -1.48   0.138    -.2749832    .0380361
vignette_name_2 |  -.0284945   .0760599    -0.37   0.708     -.177614    .1206251
vignette_name_3 |   -.097161    .076411    -1.27   0.204    -.2469688    .0526468
vignette_name_4 |  -.0811734   .0760992    -1.07   0.286      -.23037    .0680232
            age |   .0013375   .0108285     0.12   0.902    -.0198924    .0225673
          agesq |   .0057228   .0103572     0.55   0.581    -.0145831    .0260287
         female |   -.075478   .0584121    -1.29   0.196    -.1899981    .0390421
        married |  -.0068887   .0694248    -0.10   0.921    -.1429997    .1292223
        nhblack |  -.0872826   .1342889    -0.65   0.516     -.350563    .1759979
        nhother |   -.055789   .1074913    -0.52   0.604    -.2665314    .1549534
       hispanic |  -.0942861   .1215208    -0.78   0.438    -.3325341     .143962
     ed_dropout |   .1380414   .1609612     0.86   0.391    -.1775316    .4536144
     ed_hschool |   .1036221   .0854708     1.21   0.225    -.0639479    .2711921
     ed_college |    .019404   .0730489     0.27   0.791    -.1238122    .1626202
     ed_graduat |   .2235513   .0765962     2.92   0.004     .0733803    .3737223
     hinc_25_50 |    .046334   .1082473     0.43   0.669    -.1658907    .2585586
     hinc_50_75 |  -.0701819    .107471    -0.65   0.514    -.2808846    .1405207
    hinc_75_100 |  -.1042624   .1189632    -0.88   0.381    -.3374961    .1289714
     hinc_ge100 |  -.0983322   .1002532    -0.98   0.327     -.294884    .0982196
        hhsiz_2 |   .0416936   .0836537     0.50   0.618    -.1223138     .205701
        hhsiz_3 |    .252402   .1134832     2.22   0.026     .0299121    .4748919
       hhsiz_4p |   .1780106   .1346359     1.32   0.186    -.0859502    .4419714
        anykids |  -.2388215   .1008865    -2.37   0.018    -.4366148   -.0410283
          _cons |   9.345813   .3058455    30.56   0.000     8.746187     9.94544
---------------------------------------------------------------------------------

. 
. ** Row 5: P-Value on LS Starting Values
. testparm ls_startvalue_*

 ( 1)  ls_startvalue_2 = 0
 ( 2)  ls_startvalue_3 = 0

       F(  2,  4027) =   25.71
            Prob > F =    0.0000

. 
. ** Row 7: P-Value on SS Benefit Amounts
. testparm ss_benefit_*

 ( 1)  ss_benefit_2 = 0
 ( 2)  ss_benefit_3 = 0
 ( 3)  ss_benefit_4 = 0

       F(  3,  4027) =    1.06
            Prob > F =    0.3633

. 
. ** Row 8: P-Value on Vignette Names
. testparm vignette_name_*

 ( 1)  vignette_name_2 = 0
 ( 2)  vignette_name_3 = 0
 ( 3)  vignette_name_4 = 0

       F(  3,  4027) =    0.70
            Prob > F =    0.5517

. 
. 
. 
. * ------ Column 3 --------
. * Estimates for rows 1-4, row 6: the p-value on lump-sum shown first (under "P>|t|"), R2 and N are in the output below  
. reg logbuyprice  any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      10.12
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0672
                                                Root MSE          =     2.0579

---------------------------------------------------------------------------------
                |               Robust
    logbuyprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |  -.1367219    .068086    -2.01   0.045    -.2702081   -.0032356
    consequence |   .1334814   .0648055     2.06   0.039     .0064268    .2605359
     cognix_pca |   .0981797   .0456418     2.15   0.032     .0086965    .1876629
     sell_first |   .7769918   .0653356    11.89   0.000     .6488978    .9050858
ls_startvalue_2 |   .2355473   .0792337     2.97   0.003     .0802054    .3908893
ls_startvalue_3 |   .4764367   .0788874     6.04   0.000     .3217737    .6310996
       ls_first |  -.0649965   .0647657    -1.00   0.316    -.1919731    .0619801
   ss_benefit_2 |  -.4584842   .0925164    -4.96   0.000    -.6398674    -.277101
   ss_benefit_3 |  -.3934582   .0906761    -4.34   0.000    -.5712335   -.2156829
   ss_benefit_4 |  -.3534632    .092986    -3.80   0.000    -.5357672   -.1711591
vignette_name_2 |  -.0983106   .0886346    -1.11   0.267    -.2720835    .0754624
vignette_name_3 |   .1140095   .0913234     1.25   0.212    -.0650349    .2930539
vignette_name_4 |   .1462166   .0890904     1.64   0.101    -.0284498     .320883
            age |   -.034668   .0129412    -2.68   0.007    -.0600399   -.0092961
          agesq |    .023434   .0129005     1.82   0.069    -.0018581    .0487262
         female |  -.1595815   .0687841    -2.32   0.020    -.2944364   -.0247265
        married |  -.1037356    .081465    -1.27   0.203    -.2634522    .0559809
        nhblack |  -.1158509   .1477744    -0.78   0.433    -.4055704    .1738686
        nhother |   -.087201   .1283783    -0.68   0.497    -.3388934    .1644915
       hispanic |  -.0968947   .1329446    -0.73   0.466    -.3575396    .1637502
     ed_dropout |    .135633   .1818374     0.75   0.456     -.220869    .4921349
     ed_hschool |    .047632   .0989367     0.48   0.630    -.1463387    .2416028
     ed_college |   .1000541   .0864912     1.16   0.247    -.0695165    .2696246
     ed_graduat |   .2327128   .0995416     2.34   0.019     .0375562    .4278693
     hinc_25_50 |  -.1358911   .1228556    -1.11   0.269    -.3767561     .104974
     hinc_50_75 |  -.0372692    .121471    -0.31   0.759    -.2754195    .2008811
    hinc_75_100 |  -.0104659   .1316783    -0.08   0.937    -.2686282    .2476965
     hinc_ge100 |  -.0407582   .1112642    -0.37   0.714    -.2588977    .1773812
        hhsiz_2 |  -.0070515   .0996338    -0.07   0.944    -.2023889    .1882858
        hhsiz_3 |  -.0250649   .1343336    -0.19   0.852     -.288433    .2383032
       hhsiz_4p |  -.2070519   .1511576    -1.37   0.171    -.5034044    .0893007
        anykids |   .1168671   .1135367     1.03   0.303    -.1057277     .339462
          _cons |   9.608916   .3506399    27.40   0.000     8.921467    10.29636
---------------------------------------------------------------------------------

. 
. * Row 5: P-Value on LS Startinv Values
. testparm ls_startvalue_*

 ( 1)  ls_startvalue_2 = 0
 ( 2)  ls_startvalue_3 = 0

       F(  2,  4027) =   18.24
            Prob > F =    0.0000

. 
. * Row 7: P-Value on SS Benefit Amounts
. testparm ss_benefit_*

 ( 1)  ss_benefit_2 = 0
 ( 2)  ss_benefit_3 = 0
 ( 3)  ss_benefit_4 = 0

       F(  3,  4027) =    9.88
            Prob > F =    0.0000

. 
. * Row 8: P-Value on Vignette Names
. testparm vignette_name_*

 ( 1)  vignette_name_2 = 0
 ( 2)  vignette_name_3 = 0
 ( 3)  vignette_name_4 = 0

       F(  3,  4027) =    2.92
            Prob > F =    0.0327

. 
. 
.         
.         
.         
.         
.         
. 
. 
. // ***************************************************************************************
. //                      TABLE 4: HETEROGENEITY IN TREATMENT EFFECTS
. // ***************************************************************************************
.     
. ** ------- Specification 1: "By Consequence Message" ---------
. **
. ** Two Rows:   1. "No consequence message"
. **             2. "Consequence message"
. **
. ** Note: here the column with coefficients on consequence message stays empty
. ** In the row "No consequence message", the coefficient on anycompXcons0 appears in the column for "Complexity Treatment"
. ** In the row "Consequence message", the coefficient on anycompXcons1 appears in the column for "Complexity Treatment"
. 
. ** Create interaction variables of complexity with consequence dummies
. gen anycompXcons0= any_complexity*(consequence==0)

. gen anycompXcons1= any_complexity*(consequence==1)

. 
. ** Output for specification 1
. reg logspread   anycompXcons*  consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =      23.37
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1569
                                                Root MSE          =     1.9604

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
  anycompXcons0 |   .1852606   .0935987     1.98   0.048     .0017552    .3687659
  anycompXcons1 |   .0784057    .089412     0.88   0.381    -.0968914    .2537027
    consequence |  -.0708256    .103714    -0.68   0.495    -.2741625    .1325112
     cognix_pca |   -.787627   .0427575   -18.42   0.000    -.8714553   -.7037987
     sell_first |   .1664503   .0616812     2.70   0.007     .0455211    .2873795
ls_startvalue_2 |   .0645282    .076118     0.85   0.397    -.0847051    .2137616
ls_startvalue_3 |  -.0020303    .074958    -0.03   0.978    -.1489894    .1449289
       ls_first |   .0296412   .0616135     0.48   0.630    -.0911553    .1504378
   ss_benefit_2 |   .1133814   .0870459     1.30   0.193    -.0572768    .2840395
   ss_benefit_3 |   .0586582   .0842711     0.70   0.486    -.1065599    .2238763
   ss_benefit_4 |   .1694925   .0869659     1.95   0.051    -.0010088    .3399938
vignette_name_2 |   .1141448   .0856267     1.33   0.183    -.0537308    .2820205
vignette_name_3 |   .0883471   .0877673     1.01   0.314    -.0837255    .2604196
vignette_name_4 |  -.0101495   .0851066    -0.12   0.905    -.1770056    .1567065
            age |     .02465   .0129627     1.90   0.057    -.0007641    .0500642
          agesq |  -.0146048   .0128824    -1.13   0.257    -.0398615    .0106518
         female |   .0861268   .0663325     1.30   0.194    -.0439216    .2161753
        married |   .0974667   .0760271     1.28   0.200    -.0515885    .2465218
        nhblack |   .0286047   .1419565     0.20   0.840    -.2497087     .306918
        nhother |   .0497715   .1215819     0.41   0.682    -.1885963    .2881394
       hispanic |   .0833963   .1251934     0.67   0.505     -.162052    .3288446
     ed_dropout |  -.0596556   .1776992    -0.34   0.737    -.4080443    .2887331
     ed_hschool |   .0326979   .0930184     0.35   0.725    -.1496695    .2150654
     ed_college |   .0085039   .0838193     0.10   0.919    -.1558283    .1728361
     ed_graduat |   .0770202   .0954137     0.81   0.420    -.1100435    .2640839
     hinc_25_50 |   .1013805   .1166947     0.87   0.385    -.1274058    .3301667
     hinc_50_75 |  -.1674852   .1159543    -1.44   0.149    -.3948197    .0598494
    hinc_75_100 |  -.0557577   .1299173    -0.43   0.668    -.3104676    .1989521
     hinc_ge100 |  -.2568181   .1098629    -2.34   0.019    -.4722101   -.0414261
        hhsiz_2 |  -.0239415   .0954281    -0.25   0.802    -.2110333    .1631504
        hhsiz_3 |   .1472842   .1306527     1.13   0.260    -.1088674    .4034359
       hhsiz_4p |   .1816359   .1452824     1.25   0.211     -.103198    .4664698
        anykids |  -.1779497   .1056178    -1.68   0.092     -.385019    .0291196
          _cons |   1.071937   .3442355     3.11   0.002     .3970446    1.746829
---------------------------------------------------------------------------------

.  
. ** P-value in the Complexity treatment column on test of equal coefficients:
. testparm anycompXcons*, equal

 ( 1)  - anycompXcons0 + anycompXcons1 = 0

       F(  1,  4026) =    0.68
            Prob > F =    0.4081

. 
. ** P-value in the Consequence message treatment column: 
. ** Stays empty (N/A)
. 
. 
. ** --------- Specification 2: "By Complexity Treatment" ----------
. **
. ** Two Rows:   1. "No complexity treatment"
. **             2. "Complexity treatment"
. **
. ** Note: here the column with coefficients on complexity treatment stays empty
. ** Note: This is technically the exact same regression as in panel A, just a different linear combination of the 
. **       regressors (e.g., note that all other coefficients, R2, RMSE, F etc. are all the same)
. 
. ** Create interaction variables of consequence message with complexity dummies
. gen consXanycomp0= consequence*(any_complexity==0)

. gen consXanycomp1= consequence*(any_complexity==1)

. 
. ** Output for specification 2
. reg logspread   any_complexity consXanycomp*  cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =      23.37
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1569
                                                Root MSE          =     1.9604

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1852606   .0935987     1.98   0.048     .0017552    .3687659
  consXanycomp0 |  -.0708256    .103714    -0.68   0.495    -.2741625    .1325112
  consXanycomp1 |  -.1776805   .0767518    -2.32   0.021    -.3281565   -.0272046
     cognix_pca |   -.787627   .0427575   -18.42   0.000    -.8714553   -.7037987
     sell_first |   .1664503   .0616812     2.70   0.007     .0455211    .2873795
ls_startvalue_2 |   .0645282    .076118     0.85   0.397    -.0847051    .2137616
ls_startvalue_3 |  -.0020303    .074958    -0.03   0.978    -.1489894    .1449289
       ls_first |   .0296412   .0616135     0.48   0.630    -.0911553    .1504378
   ss_benefit_2 |   .1133814   .0870459     1.30   0.193    -.0572768    .2840395
   ss_benefit_3 |   .0586582   .0842711     0.70   0.486    -.1065599    .2238763
   ss_benefit_4 |   .1694925   .0869659     1.95   0.051    -.0010088    .3399938
vignette_name_2 |   .1141448   .0856267     1.33   0.183    -.0537308    .2820205
vignette_name_3 |   .0883471   .0877673     1.01   0.314    -.0837255    .2604196
vignette_name_4 |  -.0101495   .0851066    -0.12   0.905    -.1770056    .1567065
            age |     .02465   .0129627     1.90   0.057    -.0007641    .0500642
          agesq |  -.0146048   .0128824    -1.13   0.257    -.0398615    .0106518
         female |   .0861268   .0663325     1.30   0.194    -.0439216    .2161753
        married |   .0974667   .0760271     1.28   0.200    -.0515885    .2465218
        nhblack |   .0286047   .1419565     0.20   0.840    -.2497087     .306918
        nhother |   .0497715   .1215819     0.41   0.682    -.1885963    .2881394
       hispanic |   .0833963   .1251934     0.67   0.505     -.162052    .3288446
     ed_dropout |  -.0596556   .1776992    -0.34   0.737    -.4080443    .2887331
     ed_hschool |   .0326979   .0930184     0.35   0.725    -.1496695    .2150654
     ed_college |   .0085039   .0838193     0.10   0.919    -.1558283    .1728361
     ed_graduat |   .0770202   .0954137     0.81   0.420    -.1100435    .2640839
     hinc_25_50 |   .1013805   .1166947     0.87   0.385    -.1274058    .3301667
     hinc_50_75 |  -.1674852   .1159543    -1.44   0.149    -.3948197    .0598494
    hinc_75_100 |  -.0557577   .1299173    -0.43   0.668    -.3104676    .1989521
     hinc_ge100 |  -.2568181   .1098629    -2.34   0.019    -.4722101   -.0414261
        hhsiz_2 |  -.0239415   .0954281    -0.25   0.802    -.2110333    .1631504
        hhsiz_3 |   .1472842   .1306527     1.13   0.260    -.1088674    .4034359
       hhsiz_4p |   .1816359   .1452824     1.25   0.211     -.103198    .4664698
        anykids |  -.1779497   .1056178    -1.68   0.092     -.385019    .0291196
          _cons |   1.071937   .3442355     3.11   0.002     .3970446    1.746829
---------------------------------------------------------------------------------

.  
. ** P-value in the Complexity treatment column:
. ** Stays empty (N/A)
. 
. 
. ** P-value in the Consequence message treatment column on test of equal coefficients: 
. testparm consXanycomp*, equal

 ( 1)  - consXanycomp0 + consXanycomp1 = 0

       F(  1,  4026) =    0.68
            Prob > F =    0.4081

. 
. 
. ** ---------- Specification 3: "By Cognition" ------------------
. **
. ** Two Rows:   1. "Below median cognition index"
. **             2. "Above median cognition index"
. **
. 
. ** find the median level of cognition within the basesample
. sum cognix_pca if basesample, d

       Principal Comp. Cognition Score (standardized)
-------------------------------------------------------------
      Percentiles      Smallest
 1%     -2.33777      -3.721495
 5%    -1.735477      -3.649997
10%     -1.34894      -3.476389       Obs               4,060
25%     -.723738       -3.44727       Sum of Wgt.       4,060

50%     .0601967                      Mean           8.58e-11
                        Largest       Std. Dev.             1
75%     .7692014       2.088592
90%     1.286047       2.098501       Variance              1
95%     1.518391       2.171765       Skewness      -.2885747
99%     1.849152       2.184767       Kurtosis       2.557361

. assert cognix_pca<. if basesample

. 
. ** Create interaction variables
. gen cogn_abovemed = cognix_pca > r(p50)

. 
. gen anycompXcogn0= any_complexity*(cognix_pca <= r(p50))

. gen anycompXcogn1= any_complexity*(cognix_pca >  r(p50))

. 
. gen consXcogn0= consequence*(cognix_pca <= r(p50))

. gen consXcogn1= consequence*(cognix_pca >  r(p50))

. 
. 
. ** Output for specification 3
. reg logspread   anycompXcogn* consXcogn* cogn_abovemed cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(35, 4024)       =      22.32
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1574
                                                Root MSE          =     1.9604

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
  anycompXcogn0 |   .1315458   .1031244     1.28   0.202    -.0706352    .3337267
  anycompXcogn1 |    .133485   .0771083     1.73   0.084      -.01769    .2846599
     consXcogn0 |  -.1672916   .0990882    -1.69   0.091    -.3615594    .0269762
     consXcogn1 |  -.1165353   .0738442    -1.58   0.115    -.2613108    .0282403
  cogn_abovemed |  -.2074064   .1519114    -1.37   0.172    -.5052368     .090424
     cognix_pca |   -.712626   .0652859   -10.92   0.000    -.8406225   -.5846295
     sell_first |   .1650304    .061707     2.67   0.008     .0440505    .2860102
ls_startvalue_2 |   .0641637   .0761019     0.84   0.399    -.0850382    .2133657
ls_startvalue_3 |  -.0038213   .0750246    -0.05   0.959    -.1509111    .1432686
       ls_first |   .0318587   .0616216     0.52   0.605    -.0889538    .1526712
   ss_benefit_2 |    .110657   .0871226     1.27   0.204    -.0601515    .2814655
   ss_benefit_3 |   .0578759   .0842394     0.69   0.492      -.10728    .2230317
   ss_benefit_4 |   .1670514   .0869225     1.92   0.055    -.0033648    .3374676
vignette_name_2 |    .111997   .0856202     1.31   0.191     -.055866      .27986
vignette_name_3 |   .0894852   .0878191     1.02   0.308    -.0826889    .2616593
vignette_name_4 |  -.0111725   .0850615    -0.13   0.896      -.17794     .155595
            age |   .0248799   .0130068     1.91   0.056    -.0006206    .0503804
          agesq |  -.0148606   .0129366    -1.15   0.251    -.0402235    .0105023
         female |   .0835787   .0663299     1.26   0.208    -.0464647     .213622
        married |    .096587   .0759527     1.27   0.204    -.0523223    .2454962
        nhblack |   .0408173   .1418562     0.29   0.774    -.2372995    .3189341
        nhother |   .0501071   .1215572     0.41   0.680    -.1882123    .2884265
       hispanic |   .0842178   .1250972     0.67   0.501     -.161042    .3294776
     ed_dropout |  -.0421706     .17747    -0.24   0.812    -.3901102    .3057689
     ed_hschool |   .0305008   .0930592     0.33   0.743    -.1519468    .2129484
     ed_college |   .0145228    .084067     0.17   0.863    -.1502951    .1793406
     ed_graduat |   .0784532   .0956542     0.82   0.412     -.109082    .2659885
     hinc_25_50 |   .0906327   .1166286     0.78   0.437    -.1380239    .3192894
     hinc_50_75 |  -.1733942   .1160664    -1.49   0.135    -.4009486    .0541601
    hinc_75_100 |  -.0631148   .1299924    -0.49   0.627    -.3179719    .1917423
     hinc_ge100 |  -.2610362   .1099239    -2.37   0.018     -.476548   -.0455245
        hhsiz_2 |  -.0246714   .0953747    -0.26   0.796    -.2116587    .1623158
        hhsiz_3 |   .1449064   .1303413     1.11   0.266    -.1106347    .4004474
       hhsiz_4p |   .1753483   .1450068     1.21   0.227    -.1089452    .4596418
        anykids |  -.1753673   .1054238    -1.66   0.096    -.3820563    .0313216
          _cons |   1.213653   .3512685     3.46   0.001     .5249717    1.902333
---------------------------------------------------------------------------------

.  
. ** P-value in the Complexity treatment column on test of equal coefficients:
. testparm anycompXcogn*, equal

 ( 1)  - anycompXcogn0 + anycompXcogn1 = 0

       F(  1,  4024) =    0.00
            Prob > F =    0.9880

.  
. ** P-value in the Consequence message treatment column on test of equal coefficients: 
. testparm consXcogn*, equal

 ( 1)  - consXcogn0 + consXcogn1 = 0

       F(  1,  4024) =    0.17
            Prob > F =    0.6819

. 
. 
. ** ----------- Specification 4: "By level of Social Security Benefits" -------------
. **
. ** Two Rows:   1. "Below median ($800 or $1200 per month)"
. **             2. "Above median ($1600 or $2000 per month)"
. **
. 
. ** Run as a split by the median level
. 
. ** Check to see what the median level of SS Benefits is
. tab ss_benefit if basesample, m

  indicates |
 advice ssb |
 R received |      Freq.     Percent        Cum.
------------+-----------------------------------
      1 800 |      1,032       25.42       25.42
     2 1200 |        983       24.21       49.63
     3 1600 |      1,056       26.01       75.64
     4 2000 |        989       24.36      100.00
------------+-----------------------------------
      Total |      4,060      100.00

. assert ss_benefit<. if basesample

. 
. ** Create interaction variables
. gen ssb_abovemed = ss_benefit > 2

. 
. gen anycompXssb0= any_complexity*(1-ssb_abovemed)

. gen anycompXssb1= any_complexity*(ssb_abovemed)

. 
. gen consXssb0= consequence*(1-ssb_abovemed)

. gen consXssb1= consequence*(ssb_abovemed)

. 
. ** Output for specification 4
. reg logspread   anycompXssb* consXssb* ssb_abovemed cognix_pca $exp_controls $demographics if basesample, robust
note: ss_benefit_3 omitted because of collinearity

Linear regression                               Number of obs     =      4,060
                                                F(34, 4025)       =      22.63
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1568
                                                Root MSE          =     1.9608

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   anycompXssb0 |   .1233748   .0921505     1.34   0.181    -.0572911    .3040408
   anycompXssb1 |   .1391623   .0912085     1.53   0.127    -.0396568    .3179815
      consXssb0 |  -.1419865   .0870248    -1.63   0.103    -.3126033    .0286303
      consXssb1 |  -.1395973   .0876289    -1.59   0.111    -.3113984    .0322039
   ssb_abovemed |   .0457585   .1344296     0.34   0.734     -.217798     .309315
     cognix_pca |  -.7876915   .0427682   -18.42   0.000    -.8715408   -.7038421
     sell_first |   .1656974   .0616256     2.69   0.007     .0448772    .2865177
ls_startvalue_2 |   .0633864   .0761976     0.83   0.406    -.0860029    .2127758
ls_startvalue_3 |  -.0022041    .075017    -0.03   0.977    -.1492789    .1448707
       ls_first |   .0293409   .0616089     0.48   0.634    -.0914467    .1501285
   ss_benefit_2 |   .1127913    .087049     1.30   0.195     -.057873    .2834556
   ss_benefit_3 |          0  (omitted)
   ss_benefit_4 |   .1096635   .0876087     1.25   0.211    -.0620981    .2814251
vignette_name_2 |    .114377   .0857506     1.33   0.182    -.0537415    .2824956
vignette_name_3 |   .0885631   .0877679     1.01   0.313    -.0835106    .2606368
vignette_name_4 |  -.0106851   .0851041    -0.13   0.900    -.1775362     .156166
            age |   .0247532     .01299     1.91   0.057    -.0007144    .0502208
          agesq |  -.0146923   .0129164    -1.14   0.255    -.0400157     .010631
         female |   .0854409   .0663314     1.29   0.198    -.0446054    .2154872
        married |   .0966574   .0760369     1.27   0.204    -.0524171    .2457319
        nhblack |   .0280744   .1419033     0.20   0.843    -.2501346    .3062835
        nhother |   .0483849   .1216175     0.40   0.691    -.1900527    .2868225
       hispanic |   .0809646    .125117     0.65   0.518     -.164334    .3262631
     ed_dropout |  -.0573754   .1779955    -0.32   0.747    -.4063452    .2915944
     ed_hschool |   .0332075   .0931403     0.36   0.721    -.1493991    .2158141
     ed_college |   .0075149   .0839012     0.09   0.929    -.1569779    .1720076
     ed_graduat |   .0760266   .0955278     0.80   0.426    -.1112607    .2633139
     hinc_25_50 |   .1018643    .116742     0.87   0.383    -.1270147    .3307432
     hinc_50_75 |  -.1659749   .1160423    -1.43   0.153    -.3934821    .0615322
    hinc_75_100 |  -.0550137   .1300437    -0.42   0.672    -.3099714    .1999439
     hinc_ge100 |  -.2565351   .1101135    -2.33   0.020    -.4724185   -.0406516
        hhsiz_2 |  -.0249333   .0954671    -0.26   0.794    -.2121016     .162235
        hhsiz_3 |   .1466604   .1306333     1.12   0.262    -.1094532    .4027739
       hhsiz_4p |    .181598   .1452429     1.25   0.211    -.1031585    .4663544
        anykids |  -.1769194   .1055502    -1.68   0.094    -.3838562    .0300175
          _cons |   1.113111   .3485136     3.19   0.001     .4298315    1.796391
---------------------------------------------------------------------------------

.  
. * P-value in the Complexity treatment column on test of equal coefficients:
. testparm anycompXssb*, equal

 ( 1)  - anycompXssb0 + anycompXssb1 = 0

       F(  1,  4025) =    0.01
            Prob > F =    0.9030

.  
. * P-value in the Consequence message treatment column on test of equal coefficients: 
. testparm consXssb*, equal

 ( 1)  - consXssb0 + consXssb1 = 0

       F(  1,  4025) =    0.00
            Prob > F =    0.9846

. 
. * For the nobs for the rows (in square brackets in the last column, check that they add up to the N of the regression)
. tab ssb_abovemed if basesample  

ssb_aboveme |
          d |      Freq.     Percent        Cum.
------------+-----------------------------------
          0 |      2,015       49.63       49.63
          1 |      2,045       50.37      100.00
------------+-----------------------------------
      Total |      4,060      100.00

.         
.         
.         
.         
.         
.         
. 
.         
. // ***************************************************************************************
. //                      APX TABLE A01: SUMMARY STATISTICS AND COMPARISON TO THE CPS
. // ***************************************************************************************
.                 
. ** Column 1: Summary statistics on the UAS DATA in our baseline sample
. tabstat agecat_* $demographics_balance if basesample, statistics(mean sd N min max) columns(statistics)

    variable |      mean        sd         N       min       max
-------------+--------------------------------------------------
agecat_18_34 |  .2229064  .4162473      4060         0         1
agecat_35_49 |  .2960591  .4565736      4060         0         1
agecat_50_64 |  .3172414  .4654596      4060         0         1
agecat_65_~s |  .1637931  .3701333      4060         0         1
         age |  48.47833  15.45153      4060        18       106
       agesq |  25.88839  15.59995      4060      3.24    112.36
      female |  .5741379  .4945339      4060         0         1
     married |   .596798  .4906011      4060         0         1
     nhwhite |  .7554187  .4298916      4060         0         1
     nhblack |  .0810345  .2729217      4060         0         1
     nhother |  .0780788  .2683286      4060         0         1
    hispanic |   .085468  .2796113      4060         0         1
  ed_dropout |  .0534483  .2249534      4060         0         1
  ed_hschool |  .1933498  .3949735      4060         0         1
  ed_somecol |   .387931  .4873387      4060         0         1
  ed_college |   .217734  .4127564      4060         0         1
  ed_graduat |  .1475369  .3546841      4060         0         1
   hinc_lt25 |  .1655172  .3716925      4060         0         1
  hinc_25_50 |  .1758621  .3807497      4060         0         1
  hinc_50_75 |   .164532  .3708033      4060         0         1
 hinc_75_100 |   .129803  .3361279      4060         0         1
  hinc_ge100 |  .3642857  .4812886      4060         0         1
     hhsiz_1 |  .2012315  .4009701      4060         0         1
     hhsiz_2 |  .3896552  .4877321      4060         0         1
     hhsiz_3 |  .1738916  .3790629      4060         0         1
    hhsiz_4p |  .2352217  .4241895      4060         0         1
     anykids |  .3280788  .4695715      4060         0         1
----------------------------------------------------------------

. 
. 
. 
.         
. ** Column 2: Summary Statistics from the 2016 CPS
. ** 
. ** CPS abstract obtained from the IPUMScps (http://cps.ipums.org/cps/)
. **
. ** Year: 2016 Annual Social and Economic Supplement (taken in March; 2017 ASEC available, but 2016 used to match timing of UAS sample)
. ** Age selection: 18+
. 
. preserve

. clear

. 
. quietly infix              ///
>   int     year      1-4    ///
>   long    serial    5-9    ///
>   byte    numprec   10-11  ///
>   double  hwtsupp   12-21  ///
>   byte    gq        22-22  ///
>   byte    asecflag  23-23  ///
>   double  hhincome  24-31  ///
>   byte    month     32-33  ///
>   byte    pernum    34-35  ///
>   double  wtsupp    36-45  ///
>   byte    nchild    46-46  ///
>   byte    age       47-48  ///
>   byte    sex       49-49  ///
>   int     race      50-52  ///
>   byte    marst     53-53  ///
>   int     hispan    54-56  ///
>   int     educ      57-59  ///
>   using `"cps_ASEC2016.dat"'

. 
. replace hwtsupp  = hwtsupp  / 10000
(134,562 real changes made)

. replace wtsupp   = wtsupp   / 10000
(134,562 real changes made)

. 
. format hwtsupp  %10.4f

. format hhincome %8.0f

. format wtsupp   %10.4f

. 
. label var year     `"Survey year"'

. label var serial   `"Household serial number"'

. label var numprec  `"Number of person records following"'

. label var hwtsupp  `"Household weight, Supplement"'

. label var gq       `"Group Quarters status"'

. label var asecflag `"Flag for ASEC"'

. label var hhincome `"Total household income"'

. label var month    `"Month"'

. label var pernum   `"Person number in sample unit"'

. label var wtsupp   `"Supplement Weight"'

. label var nchild   `"Number of own children in household"'

. label var age      `"Age"'

. label var sex      `"Sex"'

. label var race     `"Race"'

. label var marst    `"Marital status"'

. label var hispan   `"Hispanic origin"'

. label var educ     `"Educational attainment recode"'

. 
. label define gq_lbl 0 `"NIU (Vacant units)"'

. label define gq_lbl 1 `"Households"', add

. label define gq_lbl 2 `"Group Quarters"', add

. label values gq gq_lbl

. 
. label define asecflag_lbl 1 `"ASEC"'

. label define asecflag_lbl 2 `"March Basic"', add

. label values asecflag asecflag_lbl

. 
. label define sex_lbl 1 `"Male"'

. label define sex_lbl 2 `"Female"', add

. label define sex_lbl 9 `"NIU"', add

. label values sex sex_lbl

. 
. label define race_lbl 100 `"White"'

. label define race_lbl 200 `"Black/Negro"', add

. label define race_lbl 300 `"American Indian/Aleut/Eskimo"', add

. label define race_lbl 650 `"Asian or Pacific Islander"', add

. label define race_lbl 651 `"Asian only"', add

. label define race_lbl 652 `"Hawaiian/Pacific Islander only"', add

. label define race_lbl 700 `"Other (single) race, n.e.c."', add

. label define race_lbl 801 `"White-Black"', add

. label define race_lbl 802 `"White-American Indian"', add

. label define race_lbl 803 `"White-Asian"', add

. label define race_lbl 804 `"White-Hawaiian/Pacific Islander"', add

. label define race_lbl 805 `"Black-American Indian"', add

. label define race_lbl 806 `"Black-Asian"', add

. label define race_lbl 807 `"Black-Hawaiian/Pacific Islander"', add

. label define race_lbl 808 `"American Indian-Asian"', add

. label define race_lbl 809 `"Asian-Hawaiian/Pacific Islander"', add

. label define race_lbl 810 `"White-Black-American Indian"', add

. label define race_lbl 811 `"White-Black-Asian"', add

. label define race_lbl 812 `"White-American Indian-Asian"', add

. label define race_lbl 813 `"White-Asian-Hawaiian/Pacific Islander"', add

. label define race_lbl 814 `"White-Black-American Indian-Asian"', add

. label define race_lbl 815 `"American Indian-Hawaiian/Pacific Islander"', add

. label define race_lbl 816 `"White-Black--Hawaiian/Pacific Islander"', add

. label define race_lbl 817 `"White-American Indian-Hawaiian/Pacific Islander"', add

. label define race_lbl 818 `"Black-American Indian-Asian"', add

. label define race_lbl 819 `"White-American Indian-Asian-Hawaiian/Pacific Islander"', add

. label define race_lbl 820 `"Two or three races, unspecified"', add

. label define race_lbl 830 `"Four or five races, unspecified"', add

. label define race_lbl 999 `"Blank"', add

. label values race race_lbl

. 
. label define marst_lbl 1 `"Married, spouse present"'

. label define marst_lbl 2 `"Married, spouse absent"', add

. label define marst_lbl 3 `"Separated"', add

. label define marst_lbl 4 `"Divorced"', add

. label define marst_lbl 5 `"Widowed"', add

. label define marst_lbl 6 `"Never married/single"', add

. label define marst_lbl 7 `"Widowed or Divorced"', add

. label define marst_lbl 9 `"NIU"', add

. label values marst marst_lbl

. 
. label define hispan_lbl 000 `"Not Hispanic"'

. label define hispan_lbl 100 `"Mexican"', add

. label define hispan_lbl 102 `"Mexican American"', add

. label define hispan_lbl 103 `"Mexicano/Mexicana"', add

. label define hispan_lbl 104 `"Chicano/Chicana"', add

. label define hispan_lbl 108 `"Mexican (Mexicano)"', add

. label define hispan_lbl 109 `"Mexicano/Chicano"', add

. label define hispan_lbl 200 `"Puerto Rican"', add

. label define hispan_lbl 300 `"Cuban"', add

. label define hispan_lbl 400 `"Dominican"', add

. label define hispan_lbl 500 `"Salvadoran"', add

. label define hispan_lbl 401 `"Other Hispanic"', add

. label define hispan_lbl 410 `"Central/South American"', add

. label define hispan_lbl 411 `"Central American, (excluding Salvadoran)"', add

. label define hispan_lbl 412 `"South American"', add

. label define hispan_lbl 901 `"Do not know"', add

. label define hispan_lbl 902 `"N/A (and no response 1985-87)"', add

. label values hispan hispan_lbl

. 
. label define educ_lbl 000 `"NIU or no schooling"'

. label define educ_lbl 001 `"NIU or blank"', add

. label define educ_lbl 002 `"None or preschool"', add

. label define educ_lbl 010 `"Grades 1, 2, 3, or 4"', add

. label define educ_lbl 011 `"Grade 1"', add

. label define educ_lbl 012 `"Grade 2"', add

. label define educ_lbl 013 `"Grade 3"', add

. label define educ_lbl 014 `"Grade 4"', add

. label define educ_lbl 020 `"Grades 5 or 6"', add

. label define educ_lbl 021 `"Grade 5"', add

. label define educ_lbl 022 `"Grade 6"', add

. label define educ_lbl 030 `"Grades 7 or 8"', add

. label define educ_lbl 031 `"Grade 7"', add

. label define educ_lbl 032 `"Grade 8"', add

. label define educ_lbl 040 `"Grade 9"', add

. label define educ_lbl 050 `"Grade 10"', add

. label define educ_lbl 060 `"Grade 11"', add

. label define educ_lbl 070 `"Grade 12"', add

. label define educ_lbl 071 `"12th grade, no diploma"', add

. label define educ_lbl 072 `"12th grade, diploma unclear"', add

. label define educ_lbl 073 `"High school diploma or equivalent"', add

. label define educ_lbl 080 `"1 year of college"', add

. label define educ_lbl 081 `"Some college but no degree"', add

. label define educ_lbl 090 `"2 years of college"', add

. label define educ_lbl 091 `"Associate's degree, occupational/vocational program"', add

. label define educ_lbl 092 `"Associate's degree, academic program"', add

. label define educ_lbl 100 `"3 years of college"', add

. label define educ_lbl 110 `"4 years of college"', add

. label define educ_lbl 111 `"Bachelor's degree"', add

. label define educ_lbl 120 `"5+ years of college"', add

. label define educ_lbl 121 `"5 years of college"', add

. label define educ_lbl 122 `"6+ years of college"', add

. label define educ_lbl 123 `"Master's degree"', add

. label define educ_lbl 124 `"Professional school degree"', add

. label define educ_lbl 125 `"Doctorate degree"', add

. label define educ_lbl 999 `"Missing/Unknown"', add

. label values educ educ_lbl

. 
. ** Check we only have ASEC observations
. assert asecflag==1

. 
. ** Drop institutionalized population
. drop if gq==2
(142 observations deleted)

. drop gq 

. 
. ** gender
. tab sex, m

        Sex |      Freq.     Percent        Cum.
------------+-----------------------------------
       Male |     63,892       47.53       47.53
     Female |     70,528       52.47      100.00
------------+-----------------------------------
      Total |    134,420      100.00

. gen female=sex==2 if sex<.

. 
. ** age categories
. recode age (18/34=1 "18-34") (35/49=2 "35-49") (50/64=3 "50-64") (65/85=4 "65+"), gen(agecat)
(134420 differences between age and agecat)

. tab agecat

  RECODE of |
  age (Age) |      Freq.     Percent        Cum.
------------+-----------------------------------
      18-34 |     39,352       29.28       29.28
      35-49 |     37,866       28.17       57.45
      50-64 |     33,539       24.95       82.40
        65+ |     23,663       17.60      100.00
------------+-----------------------------------
      Total |    134,420      100.00

. assert agecat==1|agecat==2|agecat==3|agecat==4

. gen agecat_18_34=agecat==1

. gen agecat_35_49=agecat==2

. gen agecat_50_64=agecat==3

. gen agecat_65_plus=agecat==4

. 
. ** race/ethnicity
. tab race, m

                                   Race |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                  White |    103,748       77.18       77.18
                            Black/Negro |     16,945       12.61       89.79
           American Indian/Aleut/Eskimo |      2,195        1.63       91.42
                             Asian only |      8,370        6.23       97.65
         Hawaiian/Pacific Islander only |        706        0.53       98.17
                            White-Black |        500        0.37       98.54
                  White-American Indian |        983        0.73       99.28
                            White-Asian |        371        0.28       99.55
        White-Hawaiian/Pacific Islander |        118        0.09       99.64
                  Black-American Indian |        116        0.09       99.73
                            Black-Asian |         30        0.02       99.75
        Black-Hawaiian/Pacific Islander |         10        0.01       99.76
                  American Indian-Asian |          7        0.01       99.76
        Asian-Hawaiian/Pacific Islander |        103        0.08       99.84
            White-Black-American Indian |         94        0.07       99.91
                      White-Black-Asian |         12        0.01       99.92
            White-American Indian-Asian |          5        0.00       99.92
  White-Asian-Hawaiian/Pacific Islander |         79        0.06       99.98
      White-Black-American Indian-Asian |          3        0.00       99.98
American Indian-Hawaiian/Pacific Island |          1        0.00       99.98
 White-Black--Hawaiian/Pacific Islander |          2        0.00       99.98
White-American Indian-Hawaiian/Pacific  |          3        0.00       99.99
            Black-American Indian-Asian |          1        0.00       99.99
White-American Indian-Asian-Hawaiian/Pa |          4        0.00       99.99
        Two or three races, unspecified |          2        0.00       99.99
        Four or five races, unspecified |         12        0.01      100.00
----------------------------------------+-----------------------------------
                                  Total |    134,420      100.00

. tab race, sum(race)

            |           Summary of Race
       Race |        Mean   Std. Dev.       Freq.
------------+------------------------------------
      White |         100           0     103,748
  Black/Neg |         200           0      16,945
  American  |         300           0       2,195
  Asian onl |         651           0       8,370
  Hawaiian/ |         652           0         706
  White-Bla |         801           0         500
  White-Ame |         802           0         983
  White-Asi |         803           0         371
  White-Haw |         804           0         118
  Black-Ame |         805           0         116
  Black-Asi |         806           0          30
  Black-Haw |         807           0          10
  American  |         808           0           7
  Asian-Haw |         809           0         103
  White-Bla |         810           0          94
  White-Bla |         811           0          12
  White-Ame |         812           0           5
  White-Asi |         813           0          79
  White-Bla |         814           0           3
  American  |         815           0           1
  White-Bla |         816           0           2
  White-Ame |         817           0           3
  Black-Ame |         818           0           1
  White-Ame |         819           0           4
  Two or th |         820           0           2
  Four or f |         830           0          12
------------+------------------------------------
      Total |   165.93467   164.66607     134,420

. 
. tab hispan, m

                        Hispanic origin |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                           Not Hispanic |    110,226       82.00       82.00
                                Mexican |     14,962       11.13       93.13
                           Puerto Rican |      2,079        1.55       94.68
                                  Cuban |        972        0.72       95.40
                              Dominican |        726        0.54       95.94
                         Other Hispanic |      1,671        1.24       97.18
Central American, (excluding Salvadoran |      1,367        1.02       98.20
                         South American |      1,543        1.15       99.35
                             Salvadoran |        874        0.65      100.00
----------------------------------------+-----------------------------------
                                  Total |    134,420      100.00

. tab hispan, sum(hispan)

   Hispanic |     Summary of Hispanic origin
     origin |        Mean   Std. Dev.       Freq.
------------+------------------------------------
  Not Hispa |           0           0     110,226
    Mexican |         100           0      14,962
  Puerto Ri |         200           0       2,079
      Cuban |         300           0         972
  Dominican |         400           0         726
  Other His |         401           0       1,671
  Central A |         411           0       1,367
  South Ame |         412           0       1,543
  Salvadora |         500           0         874
------------+------------------------------------
      Total |   35.698735   96.245271     134,420

. 
. gen race4=race

. recode race4 100=1 200=2 300/830=4
(race4: 134420 changes made)

. replace race4=3 if hispan>0 & hispan<.
(24,194 real changes made)

. 
. tab race4

      race4 |      Freq.     Percent        Cum.
------------+-----------------------------------
          1 |     81,363       60.53       60.53
          2 |     16,341       12.16       72.69
          3 |     24,194       18.00       90.68
          4 |     12,522        9.32      100.00
------------+-----------------------------------
      Total |    134,420      100.00

. label def race4 1 "Non-H Wh" 2 "Non-H Bl" 3 "Hispanic" 4 "Other" 

. label val race4 race4

. 
. gen byte white=race4==1

. gen byte black=race4==2

. gen byte hisp =race4==3

. gen byte other=race4==4

. label var white "Non-Hispanic White"

. label var black "Non-Hispanic Black"

. label var hisp  "Hispanic"

. label var other "Panelist is not Non-Hispanic white or black"

. assert (white+black+hisp+other)==1

. 
. ** education
. tab educ, m

          Educational attainment recode |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                      None or preschool |        449        0.33        0.33
                   Grades 1, 2, 3, or 4 |      1,022        0.76        1.09
                          Grades 5 or 6 |      2,109        1.57        2.66
                          Grades 7 or 8 |      2,329        1.73        4.40
                                Grade 9 |      2,125        1.58        5.98
                               Grade 10 |      2,538        1.89        7.86
                               Grade 11 |      4,056        3.02       10.88
                 12th grade, no diploma |      2,278        1.69       12.58
      High school diploma or equivalent |     38,932       28.96       41.54
             Some college but no degree |     25,310       18.83       60.37
Associate's degree, occupational/vocati |      5,613        4.18       64.54
   Associate's degree, academic program |      7,381        5.49       70.04
                      Bachelor's degree |     25,447       18.93       88.97
                        Master's degree |     10,845        8.07       97.03
             Professional school degree |      1,795        1.34       98.37
                       Doctorate degree |      2,191        1.63      100.00
----------------------------------------+-----------------------------------
                                  Total |    134,420      100.00

. tab educ, sum(educ)

Educational |  Summary of Educational attainment
 attainment |               recode
     recode |        Mean   Std. Dev.       Freq.
------------+------------------------------------
  None or p |           2           0         449
  Grades 1, |          10           0       1,022
  Grades 5  |          20           0       2,109
  Grades 7  |          30           0       2,329
    Grade 9 |          40           0       2,125
   Grade 10 |          50           0       2,538
   Grade 11 |          60           0       4,056
  12th grad |          71           0       2,278
  High scho |          73           0      38,932
  Some coll |          81           0      25,310
  Associate |          91           0       5,613
  Associate |          92           0       7,381
  Bachelor' |         111           0      25,447
  Master's  |         123           0      10,845
  Professio |         124           0       1,795
  Doctorate |         125           0       2,191
------------+------------------------------------
      Total |   85.382785   24.524231     134,420

. 
. gen educ5=educ

. recode educ5 2/71=1 73=2 80/110=3 111=4 120/125=5
(educ5: 134420 changes made)

. label def educ5 1 "HS dropout" 2 "HS" 3 "Some College"  4 "Bachelor's degree" 5 "Graduate degree"

. label val educ5 educ5

. 
. tab educ5

            educ5 |      Freq.     Percent        Cum.
------------------+-----------------------------------
       HS dropout |     16,906       12.58       12.58
               HS |     38,932       28.96       41.54
     Some College |     38,304       28.50       70.04
Bachelor's degree |     25,447       18.93       88.97
  Graduate degree |     14,831       11.03      100.00
------------------+-----------------------------------
            Total |    134,420      100.00

. 
. gen byte edudo  = educ5==1

. gen byte eduhs  = educ5==2

. gen byte edusc  = educ5==3

. gen byte edubd  = educ5==4

. gen byte edugd  = educ5==5

. assert (edudo+eduhs+edusc+edubd+edugd)==1

. 
. label var edudo "High School Dropout"

. label var eduhs "High School Education"

. label var edusc "Some college"

. label var edubd "Bachelor's degree"

. label var edugd "Graduate degree"

. 
. ** marital status
. tab marst, m

         Marital status |      Freq.     Percent        Cum.
------------------------+-----------------------------------
Married, spouse present |     72,582       54.00       54.00
 Married, spouse absent |      1,876        1.40       55.39
              Separated |      2,905        2.16       57.55
               Divorced |     13,788       10.26       67.81
                Widowed |      7,539        5.61       73.42
   Never married/single |     35,730       26.58      100.00
------------------------+-----------------------------------
                  Total |    134,420      100.00

. tab marst, sum(marst)

    Marital |      Summary of Marital status
     status |        Mean   Std. Dev.       Freq.
------------+------------------------------------
  Married,  |           1           0      72,582
  Married,  |           2           0       1,876
  Separated |           3           0       2,905
   Divorced |           4           0      13,788
    Widowed |           5           0       7,539
  Never mar |           6           0      35,730
------------+------------------------------------
      Total |    2.918286   2.2105126     134,420

. 
. gen byte xmarried  = marst==1|marst==2

. gen byte xsingle   = marst==3|marst==4|marst==5|marst==6

. assert (xmarried+xsingle)==1

. 
. label var xmarried "Married"

. label var xsingle "Single"

.      
.           
. ** household size
. tab numprec, m 

  Number of |
     person |
    records |
  following |      Freq.     Percent        Cum.
------------+-----------------------------------
          1 |     16,958       12.62       12.62
          2 |     40,615       30.21       42.83
          3 |     26,506       19.72       62.55
          4 |     26,246       19.53       82.07
          5 |     13,901       10.34       92.42
          6 |      5,944        4.42       96.84
          7 |      2,344        1.74       98.58
          8 |        981        0.73       99.31
          9 |        454        0.34       99.65
         10 |        213        0.16       99.81
         11 |         70        0.05       99.86
         12 |         64        0.05       99.91
         13 |         51        0.04       99.95
         14 |         21        0.02       99.96
         15 |         22        0.02       99.98
         16 |         30        0.02      100.00
------------+-----------------------------------
      Total |    134,420      100.00

. gen pphhsize=numprec

. 
. gen hhsize_1=pphhsize==1

. gen hhsize_2=pphhsize==2

. gen hhsize_3=pphhsize==3

. gen hhsize_4p=pphhsize>=4

. assert (hhsize_1+hhsize_2+hhsize_3+hhsize_4p)==1

. 
. label var hhsize_1 "Household size of one"

. label var hhsize_2 "Household size of two"

. label var hhsize_3 "Household size of three"

. label var hhsize_4 "Household size of four or more"

. 
.           
. ** household income
. gen hhinc5=hhincome

. recode hhinc5 min/24999=1 25000/49999=2 50000/74999=3 75000/99999=4 100000/max=5
(hhinc5: 134395 changes made)

. tab hhinc5

     hhinc5 |      Freq.     Percent        Cum.
------------+-----------------------------------
          1 |     21,222       15.79       15.79
          2 |     27,494       20.45       36.24
          3 |     23,455       17.45       53.69
          4 |     18,613       13.85       67.54
          5 |     43,636       32.46      100.00
------------+-----------------------------------
      Total |    134,420      100.00

. 
. gen inc00_25  = hhinc5==1

. gen inc25_50  = hhinc5==2

. gen inc50_75  = hhinc5==3

. gen inc75_100 = hhinc5==4

. gen inc100p   = hhinc5==5

. assert (inc00_25+inc25_50+inc50_75+inc75_100+inc100p)==1

. 
. label var inc00_25 "Household income: Below 25k"

. label var inc25_50 "Household income: 25k-50k"

. label var inc50_75 "Household income: 50k-75k"

. label var inc75_100 "Household income: 75k-100k"

. label var inc100p "Household income: Above 100k"

. 
. ** children present in household
. gen nchild2=nchild

. recode nchild2 1/9=1
(nchild2: 35002 changes made)

. gen nokids = nchild2==0

. gen anykids = nchild2==1

. assert (nokids+anykids)==1

. 
. sum agecat_*  female  xmarried  white  black  other  hisp  edudo  eduhs  edusc  edubd edugd inc00_25-inc100p hhsize_* anykids [aw=wtsupp],
>  sep(0)

    Variable |     Obs      Weight        Mean   Std. Dev.       Min        Max
-------------+-----------------------------------------------------------------
agecat_18_34 | 134,420   244543842    .3002277   .4583586          0          1
agecat_35_49 | 134,420   244543842    .2481045   .4319144          0          1
agecat_50_64 | 134,420   244543842    .2575811   .4373036          0          1
agecat_65_~s | 134,420   244543842    .1940867   .3954974          0          1
      female | 134,420   244543842    .5160402   .4997445          0          1
    xmarried | 134,420   244543842    .5265477   .4992966          0          1
       white | 134,420   244543842    .6436767   .4789141          0          1
       black | 134,420   244543842    .1183906   .3230713          0          1
       other | 134,420   244543842    .0803776   .2718779          0          1
        hisp | 134,420   244543842     .157555   .3643246          0          1
       edudo | 134,420   244543842    .1169759   .3213928          0          1
       eduhs | 134,420   244543842    .2895621   .4535608          0          1
       edusc | 134,420   244543842    .2863715   .4520668          0          1
       edubd | 134,420   244543842    .1950567   .3962458          0          1
       edugd | 134,420   244543842    .1120338   .3154092          0          1
    inc00_25 | 134,420   244543842    .1605378   .3671054          0          1
    inc25_50 | 134,420   244543842    .2051056   .4037803          0          1
    inc50_75 | 134,420   244543842    .1729084   .3781695          0          1
   inc75_100 | 134,420   244543842    .1375243   .3444013          0          1
     inc100p | 134,420   244543842    .3239238   .4679731          0          1
    hhsize_1 | 134,420   244543842    .1446886   .3517879          0          1
    hhsize_2 | 134,420   244543842    .3417694   .4743045          0          1
    hhsize_3 | 134,420   244543842    .1912084   .3932543          0          1
   hhsize_4p | 134,420   244543842    .3223336   .4673717          0          1
     anykids | 134,420   244543842    .3783964    .484989          0          1

. 
. restore

. 
. 
. 
. 
. 
. 
.         
. // ***************************************************************************************
. //                      APX TABLE A02: TEXT OF THE VIGNETTES AND THE CONSEQUENCE MESSAGE
. //                                              Not Part of STATA file
. // ***************************************************************************************
. 
. 
. 
. 
. 
. 
. 
. 
. // ***************************************************************************************
. //                      APX TABLE A03: BALANCE TESTS
. // ***************************************************************************************
. 
.  
. ** Given that we look at complexity and consequence uninteracted, we focus on
. ** the two pairwise balance tests 
. 
. ** Columns 1-3 of Randomization check table
. ** ----------------------------------------
. ** Col. 1: means for variables when any_complexity==0
. ** Col. 2: means for variables when any_complexity==1
. ** Col. 3: p-value on test of equal mean
. 
. ** Col 1 & 2: baseline sample demographics and cognition
. tabstat $demographics_balance cognix_pca if basesample, by(any_complexity) statistics(mean semean N) columns(statistics)

Summary for variables: age agesq female married nhwhite nhblack nhother hispanic ed_dropout ed_hschool ed_somecol ed_college ed_graduat hinc
> _lt25 hinc_25_50 hinc_50_75 hinc_75_100 hinc_ge100 hhsiz_1 hhsiz_2 hhsiz_3 hhsiz_4p anykids cognix_pca
     by categories of: any_complexity (Any Complexity)

any_complexity |      mean  se(mean)         N
---------------+------------------------------
             0 |  48.42583   .422238      1409
               |  25.96087  .4265794      1409
               |  .5798439  .0131541      1409
               |  .5734564  .0131805      1409
               |   .761533  .0113568      1409
               |  .0723918   .006906      1409
               |  .0809084  .0072673      1409
               |  .0851668  .0074388      1409
               |  .0546487  .0060574      1409
               |  .1859475  .0103686      1409
               |  .4059617  .0130873      1409
               |    .20511  .0107608      1409
               |  .1483322  .0094722      1409
               |  .1660752  .0099178      1409
               |  .1753016   .010133      1409
               |  .1497516  .0095095      1409
               |  .1405252  .0092617      1409
               |  .3683463  .0128548      1409
               |  .2150461  .0109493      1409
               |  .3775727  .0129194      1409
               |  .1767211  .0101652      1409
               |    .23066  .0112265      1409
               |  .3243435  .0124757      1409
               |  -.038991  .0269863      1409
---------------+------------------------------
             1 |  48.50622  .2959685      2651
               |  25.84987  .2986953      2651
               |  .5711052  .0096141      2651
               |  .6092041  .0094784      2651
               |   .752169  .0083871      2651
               |  .0856281  .0054356      2651
               |  .0765749  .0051656      2651
               |  .0856281  .0054356      2651
               |  .0528103  .0043446      2651
               |   .197284  .0077304      2651
               |  .3783478   .009421      2651
               |  .2244436  .0081047      2651
               |  .1471143   .006881      2651
               |  .1652207  .0072143      2651
               |  .1761599  .0074003      2651
               |  .1723878  .0073374      2651
               |  .1241041  .0064047      2651
               |  .3621275  .0093363      2651
               |  .1938891  .0076798      2651
               |   .396077  .0095007      2651
               |  .1723878  .0073374      2651
               |  .2376462  .0082684      2651
               |  .3300641  .0091347      2651
               |  .0207236  .0192784      2651
---------------+------------------------------
         Total |  48.47833  .2424981      4060
               |  25.88839  .2448276      4060
               |  .5741379  .0077613      4060
               |   .596798  .0076996      4060
               |  .7554187  .0067468      4060
               |  .0810345  .0042833      4060
               |  .0780788  .0042112      4060
               |   .085468  .0043883      4060
               |  .0534483  .0035304      4060
               |  .1933498  .0061988      4060
               |   .387931  .0076484      4060
               |   .217734  .0064778      4060
               |  .1475369  .0055665      4060
               |  .1655172  .0058334      4060
               |  .1758621  .0059755      4060
               |   .164532  .0058194      4060
               |   .129803  .0052752      4060
               |  .3642857  .0075534      4060
               |  .2012315  .0062929      4060
               |  .3896552  .0076545      4060
               |  .1738916  .0059491      4060
               |  .2352217  .0066573      4060
               |  .3280788  .0073695      4060
               |  8.58e-11  .0156941      4060
----------------------------------------------

. 
. ** append to col 1 & 2: indicators of missing data or attrition
. tabstat attrit basesample miss_spread miss_anydemographic miss_cognix_pca, by(any_complexity) statistics(mean semean N) columns(statistics
> )

Summary for variables: attrit basesample miss_spread miss_anydemographic miss_cognix_pca
     by categories of: any_complexity (Any Complexity)

any_complexity |      mean  se(mean)         N
---------------+------------------------------
             0 |  .0108905  .0026277      1561
               |  .9026265  .0075061      1561
               |  .0076874  .0022113      1561
               |  .0032031  .0014306      1561
               |  .0903267  .0072575      1561
---------------+------------------------------
             1 |  .0151565  .0022181      3035
               |  .8734761  .0060354      3035
               |  .0105437  .0018543      3035
               |  .0065898  .0014689      3035
               |  .1149918  .0057916      3035
---------------+------------------------------
         Total |  .0137076  .0017153      4596
               |  .8833768   .004735      4596
               |  .0095735  .0014365      4596
               |  .0054395  .0010851      4596
               |  .1066144  .0045529      4596
----------------------------------------------

.   
. * Col 3: test equality of means in each row
. foreach var of varlist $demographics_balance cognix_pca {
  2.    qui reg `var' any_complexity if basesample, robust
  3.    di "P-value on test of equal means = "  Ftail(e(df_m),e(df_r),e(F))   "               for `var' "
  4. }
P-value on test of equal means = .8761096               for age 
P-value on test of equal means = .83121419               for agesq 
P-value on test of equal means = .5917322               for female 
P-value on test of equal means = .02771758               for married 
P-value on test of equal means = .5071794               for nhwhite 
P-value on test of equal means = .13210502               for nhblack 
P-value on test of equal means = .62694948               for nhother 
P-value on test of equal means = .96006893               for hispanic 
P-value on test of equal means = .80520217               for ed_dropout 
P-value on test of equal means = .38076068               for ed_hschool 
P-value on test of equal means = .08687608               for ed_somecol 
P-value on test of equal means = .15130255               for ed_college 
P-value on test of equal means = .91715322               for ed_graduat 
P-value on test of equal means = .94444928               for hinc_lt25 
P-value on test of equal means = .94546476               for hinc_25_50 
P-value on test of equal means = .05954425               for hinc_50_75 
P-value on test of equal means = .14481673               for hinc_75_100 
P-value on test of equal means = .69548537               for hinc_ge100 
P-value on test of equal means = .11372201               for hhsiz_1 
P-value on test of equal means = .24859926               for hhsiz_2 
P-value on test of equal means = .72961124               for hhsiz_3 
P-value on test of equal means = .61633971               for hhsiz_4p 
P-value on test of equal means = .71140979               for anykids 
P-value on test of equal means = .07183614               for cognix_pca 

. 
. * Col 3, continued: test equality of means in each row of indicators of missing data or attrition 
. foreach var of varlist attrit basesample miss_spread miss_anydemographic miss_cognix_pca {
  2.    qui reg `var' any_complexity, robust
  3.    di "P-value on test of equal means = "  Ftail(e(df_m),e(df_r),e(F))   "               for `var' "
  4. }
P-value on test of equal means = .21480347               for attrit 
P-value on test of equal means = .00248625               for basesample 
P-value on test of equal means = .32233791               for miss_spread 
P-value on test of equal means = .098658               for miss_anydemographic 
P-value on test of equal means = .00792286               for miss_cognix_pca 

. 
. * Joint test (only variable defined in the baseline sample
. * Report in the last row of the table
. logit any_complexity $demographics cognix_pca if basesample, vce(robust)   

Iteration 0:   log pseudolikelihood = -2621.1268  
Iteration 1:   log pseudolikelihood = -2606.7978  
Iteration 2:   log pseudolikelihood = -2606.7847  
Iteration 3:   log pseudolikelihood = -2606.7847  

Logistic regression                             Number of obs     =      4,060
                                                Wald chi2(20)     =      28.12
                                                Prob > chi2       =     0.1066
Log pseudolikelihood = -2606.7847               Pseudo R2         =     0.0055

--------------------------------------------------------------------------------
               |               Robust
any_complexity |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
           age |   .0230857   .0124495     1.85   0.064    -.0013149    .0474862
         agesq |  -.0233136   .0123015    -1.90   0.058    -.0474242     .000797
        female |   .0225035   .0713672     0.32   0.753    -.1173736    .1623806
       married |   .1218695   .0802649     1.52   0.129    -.0354469    .2791859
       nhblack |   .3143899   .1358553     2.31   0.021     .0481184    .5806614
       nhother |   .0040008    .125204     0.03   0.975    -.2413946    .2493962
      hispanic |   .1164055   .1264913     0.92   0.357    -.1315128    .3643239
    ed_dropout |   .0972026   .1601214     0.61   0.544    -.2166295    .4110347
    ed_hschool |   .1646498   .0949446     1.73   0.083    -.0214381    .3507377
    ed_college |   .1261237   .0945341     1.33   0.182    -.0591597     .311407
    ed_graduat |   .0200752   .1082909     0.19   0.853     -.192171    .2323214
    hinc_25_50 |  -.0226598   .1160503    -0.20   0.845    -.2501142    .2047947
    hinc_50_75 |   .0757615   .1224335     0.62   0.536    -.1642038    .3157268
   hinc_75_100 |  -.2157298    .128868    -1.67   0.094    -.4683063    .0368468
    hinc_ge100 |  -.1377922   .1119962    -1.23   0.219    -.3573006    .0817162
       hhsiz_2 |   .0997729    .099071     1.01   0.314    -.0944027    .2939485
       hhsiz_3 |   .0068704   .1312493     0.05   0.958    -.2503734    .2641143
      hhsiz_4p |   .0350411   .1493369     0.23   0.814    -.2576538     .327736
       anykids |   .0247685   .1137195     0.22   0.828    -.1981175    .2476546
    cognix_pca |   .1015023   .0447621     2.27   0.023     .0137702    .1892343
         _cons |  -.0544058   .3202622    -0.17   0.865    -.6821081    .5732966
--------------------------------------------------------------------------------

. 
. 
. 
. ** Columns 4-6 of Randomization check table
. ** ----------------------------------------
. ** Col. 4: means for variables when consequence message ==0
. ** Col. 5: means for variables when consequence message ==1
. ** Col. 6: p-value on test of equal mean
. 
. 
. ** Col 4 & 5: baseline sample demographics and cognition
. tabstat $demographics_balance cognix_pca if basesample, by(consequence) statistics(mean semean N) columns(statistics)

Summary for variables: age agesq female married nhwhite nhblack nhother hispanic ed_dropout ed_hschool ed_somecol ed_college ed_graduat hinc
> _lt25 hinc_25_50 hinc_50_75 hinc_75_100 hinc_ge100 hhsiz_1 hhsiz_2 hhsiz_3 hhsiz_4p anykids cognix_pca
     by categories of: consequence (Consequence Treatment)

consequence |      mean  se(mean)         N
------------+------------------------------
         No |   48.5015  .3432999      1998
            |  25.87752  .3486173      1998
            |  .5775776  .0110532      1998
            |  .5930931  .0109931      1998
            |  .7587588  .0095739      1998
            |  .0825826  .0061594      1998
            |  .0740741  .0058605      1998
            |  .0845846  .0062268      1998
            |  .0520521  .0049708      1998
            |  .1961962  .0088865      1998
            |  .3783784  .0108527      1998
            |  .2177177  .0092351      1998
            |  .1556557  .0081125      1998
            |  .1561562  .0081231      1998
            |  .1806807  .0086098      1998
            |  .1651652  .0083094      1998
            |  .1251251  .0074038      1998
            |  .3728729   .010821      1998
            |  .2027027   .008996      1998
            |  .3843844  .0108855      1998
            |  .1846847  .0086834      1998
            |  .2282282  .0093916      1998
            |  .3303303  .0105248      1998
            | -.0060637  .0222225      1998
------------+------------------------------
        Yes |  48.45587  .3426084      2062
            |  25.89892  .3439891      2062
            |   .570805  .0109027      2062
            |   .600388  .0107894      2062
            |  .7521823  .0095102      2062
            |  .0795344  .0059599      2062
            |  .0819593  .0060421      2062
            |   .086324  .0061862      2062
            |  .0548012  .0050132      2062
            |  .1905917  .0086516      2062
            |  .3971872  .0107783      2062
            |  .2177498   .009091      2062
            |  .1396702  .0076356      2062
            |  .1745878  .0083619      2062
            |   .171193  .0082972      2062
            |  .1639185  .0081545      2062
            |  .1343356  .0075116      2062
            |  .3559651  .0105468      2062
            |   .199806  .0088077      2062
            |  .3947624  .0107669      2062
            |  .1634336  .0081448      2062
            |  .2419981  .0094341      2062
            |  .3258972  .0103244      2062
            |  .0058755   .022168      2062
------------+------------------------------
      Total |  48.47833  .2424981      4060
            |  25.88839  .2448276      4060
            |  .5741379  .0077613      4060
            |   .596798  .0076996      4060
            |  .7554187  .0067468      4060
            |  .0810345  .0042833      4060
            |  .0780788  .0042112      4060
            |   .085468  .0043883      4060
            |  .0534483  .0035304      4060
            |  .1933498  .0061988      4060
            |   .387931  .0076484      4060
            |   .217734  .0064778      4060
            |  .1475369  .0055665      4060
            |  .1655172  .0058334      4060
            |  .1758621  .0059755      4060
            |   .164532  .0058194      4060
            |   .129803  .0052752      4060
            |  .3642857  .0075534      4060
            |  .2012315  .0062929      4060
            |  .3896552  .0076545      4060
            |  .1738916  .0059491      4060
            |  .2352217  .0066573      4060
            |  .3280788  .0073695      4060
            |  8.58e-11  .0156941      4060
-------------------------------------------

. 
. ** append to col 4 & 5: indicators of missing data or attrition
. tabstat attrit basesample miss_spread miss_anydemographic miss_cognix_pca, by(consequence) statistics(mean semean N) columns(statistics) 

Summary for variables: attrit basesample miss_spread miss_anydemographic miss_cognix_pca
     by categories of: consequence (Consequence Treatment)

consequence |      mean  se(mean)         N
------------+------------------------------
         No |  .0154253  .0025877      2269
            |  .8805641  .0068097      2269
            |  .0105773  .0021481      2269
            |  .0061701  .0016443      2269
            |  .1088585  .0065401      2269
------------+------------------------------
        Yes |  .0120327  .0022607      2327
            |  .8861195  .0065867      2327
            |  .0085948   .001914      2327
            |  .0047271  .0014222      2327
            |  .1044263  .0063409      2327
------------+------------------------------
      Total |  .0137076  .0017153      4596
            |  .8833768   .004735      4596
            |  .0095735  .0014365      4596
            |  .0054395  .0010851      4596
            |  .1066144  .0045529      4596
-------------------------------------------

. 
. ** Col 6: test equality of means in each row
. foreach var of varlist $demographics_balance cognix_pca {
  2.    qui reg `var' consequence if basesample, robust
  3.    di "P-value on test of equal means = "  Ftail(e(df_m),e(df_r),e(F))   "               for `var' "
  4. }
P-value on test of equal means = .92504433               for age 
P-value on test of equal means = .96514006               for agesq 
P-value on test of equal means = .66270081               for female 
P-value on test of equal means = .63581422               for married 
P-value on test of equal means = .62604494               for nhwhite 
P-value on test of equal means = .72212704               for nhblack 
P-value on test of equal means = .34893117               for nhother 
P-value on test of equal means = .84292523               for hispanic 
P-value on test of equal means = .69699818               for ed_dropout 
P-value on test of equal means = .65137224               for ed_hschool 
P-value on test of equal means = .21888211               for ed_somecol 
P-value on test of equal means = .99802742               for ed_college 
P-value on test of equal means = .15140168               for ed_graduat 
P-value on test of equal means = .11394507               for hinc_lt25 
P-value on test of equal means = .4275469               for hinc_25_50 
P-value on test of equal means = .91473193               for hinc_50_75 
P-value on test of equal means = .38256554               for hinc_75_100 
P-value on test of equal means = .26323274               for hinc_ge100 
P-value on test of equal means = .81804014               for hhsiz_1 
P-value on test of equal means = .49792415               for hhsiz_2 
P-value on test of equal means = .07433699               for hhsiz_3 
P-value on test of equal means = .3010088               for hhsiz_4p 
P-value on test of equal means = .76366749               for anykids 
P-value on test of equal means = .70369366               for cognix_pca 

. 
. ** Col 6, continued: test equality of means in each row of indicators of missing data or attrition 
. foreach var of varlist attrit basesample miss_spread miss_anydemographic miss_cognix_pca {
  2.    qui reg `var' consequence, robust
  3.    di "P-value on test of equal means = "  Ftail(e(df_m),e(df_r),e(F))   "               for `var' "
  4. }
P-value on test of equal means = .32353151               for attrit 
P-value on test of equal means = .55764884               for basesample 
P-value on test of equal means = .4907999               for miss_spread 
P-value on test of equal means = .50688664               for miss_anydemographic 
P-value on test of equal means = .62659338               for miss_cognix_pca 

. 
. ** Joint test (only variable defined in the baseline sample
. ** Report in the last row of the table
. logit consequence $demographics cognix_pca if basesample, vce(robust)   

Iteration 0:   log pseudolikelihood = -2813.6731  
Iteration 1:   log pseudolikelihood = -2806.3164  
Iteration 2:   log pseudolikelihood = -2806.3163  

Logistic regression                             Number of obs     =      4,060
                                                Wald chi2(20)     =      14.80
                                                Prob > chi2       =     0.7876
Log pseudolikelihood = -2806.3163               Pseudo R2         =     0.0026

------------------------------------------------------------------------------
             |               Robust
 consequence |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         age |  -.0095011       .012    -0.79   0.429    -.0330207    .0140185
       agesq |   .0091291   .0118413     0.77   0.441    -.0140794    .0323376
      female |  -.0072808    .068036    -0.11   0.915    -.1406288    .1260673
     married |   .0544798   .0762242     0.71   0.475    -.0949168    .2038764
     nhblack |   .0046315   .1267758     0.04   0.971    -.2438446    .2531076
     nhother |   .1253036   .1205095     1.04   0.298    -.1108907    .3614978
    hispanic |   .0242414   .1205116     0.20   0.841    -.2119571    .2604398
  ed_dropout |  -.0132769   .1526883    -0.09   0.931    -.3125404    .2859866
  ed_hschool |  -.0817428   .0899941    -0.91   0.364     -.258128    .0946424
  ed_college |   -.073551   .0893352    -0.82   0.410    -.2486447    .1015428
  ed_graduat |  -.1891936   .1031761    -1.83   0.067     -.391415    .0130279
  hinc_25_50 |  -.1853204    .110087    -1.68   0.092    -.4010869    .0304461
  hinc_50_75 |  -.1450655    .114601    -1.27   0.206    -.3696794    .0795483
 hinc_75_100 |   -.062489   .1230235    -0.51   0.611    -.3036105    .1786326
  hinc_ge100 |  -.1794637    .105999    -1.69   0.090     -.387218    .0282905
     hhsiz_2 |   .0218373   .0950271     0.23   0.818    -.1644123    .2080869
     hhsiz_3 |  -.1010691   .1259357    -0.80   0.422    -.3478984    .1457603
    hhsiz_4p |   .0995512   .1432167     0.70   0.487    -.1811485    .3802509
     anykids |  -.0625264   .1087217    -0.58   0.565     -.275617    .1505642
  cognix_pca |   .0477104   .0426111     1.12   0.263    -.0358058    .1312266
       _cons |   .4118852   .3081902     1.34   0.181    -.1921566    1.015927
------------------------------------------------------------------------------

. 
. 
. 
. 
. 
.                 
. // ***************************************************************************************
. //                      APX TABLE A04: Predictors of Heterogeneity in the Sell-Buy Spread in Control Sample
. // ***************************************************************************************
. 
. 
. ** -------- Description of heterogeneity in spread in control sample ------- 
. ** Show to what extent demographics and the cognition index can explain the spread 
. ** Exclude the other experimental controls so that R2 shows variation explained by
. ** demographics and/or cognition index
. 
. ** Column 1: Just demographics. Limit to sample to exclude complexity tretments or consequence message treatment
. reg logspread                 $demographics if basesample & !any_complexity & !consequence, robust

Linear regression                               Number of obs     =        708
                                                F(19, 688)        =       4.35
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1080
                                                Root MSE          =      2.051

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         age |   .0167298   .0272029     0.61   0.539    -.0366809    .0701404
       agesq |  -.0232989    .025146    -0.93   0.354    -.0726709    .0260732
      female |   .5039445   .1644289     3.06   0.002     .1811018    .8267873
     married |  -.0732531   .1891721    -0.39   0.699     -.444677    .2981708
     nhblack |   .8468007   .3753012     2.26   0.024     .1099275    1.583674
     nhother |   .4534303   .3013626     1.50   0.133    -.1382705    1.045131
    hispanic |  -.0885242   .2922787    -0.30   0.762    -.6623895    .4853411
  ed_dropout |   1.045945   .4583237     2.28   0.023     .1460641    1.945826
  ed_hschool |   .5821708   .2381672     2.44   0.015     .1145491    1.049792
  ed_college |  -.4630369   .1990407    -2.33   0.020    -.8538369   -.0722368
  ed_graduat |  -.2685959   .2208903    -1.22   0.224    -.7022959    .1651042
  hinc_25_50 |   .1569637   .2948234     0.53   0.595    -.4218978    .7358252
  hinc_50_75 |   .1657101   .2971057     0.56   0.577    -.4176326    .7490527
 hinc_75_100 |  -.3694922   .3084803    -1.20   0.231     -.975168    .2361835
  hinc_ge100 |   -.317842   .2596392    -1.22   0.221    -.8276224    .1919384
     hhsiz_2 |  -.2444331   .2418062    -1.01   0.312    -.7191997    .2303335
     hhsiz_3 |   .0080176   .3341663     0.02   0.981    -.6480907    .6641258
    hhsiz_4p |  -.2433608   .3791891    -0.64   0.521    -.9878675    .5011458
     anykids |  -.1358819   .2897875    -0.47   0.639    -.7048559    .4330921
       _cons |   1.943855   .7560365     2.57   0.010     .4594396    3.428271
------------------------------------------------------------------------------

. 
. ** Rather than reporting each of these dummies, report p-values on their joint tests of being zero
. testparm nhblack nhother hispanic

 ( 1)  nhblack = 0
 ( 2)  nhother = 0
 ( 3)  hispanic = 0

       F(  3,   688) =    2.45
            Prob > F =    0.0625

. testparm ed_*

 ( 1)  ed_dropout = 0
 ( 2)  ed_hschool = 0
 ( 3)  ed_college = 0
 ( 4)  ed_graduat = 0

       F(  4,   688) =    5.74
            Prob > F =    0.0002

. testparm hinc_*

 ( 1)  hinc_25_50 = 0
 ( 2)  hinc_50_75 = 0
 ( 3)  hinc_75_100 = 0
 ( 4)  hinc_ge100 = 0

       F(  4,   688) =    1.87
            Prob > F =    0.1140

. testparm hhsiz_*

 ( 1)  hhsiz_2 = 0
 ( 2)  hhsiz_3 = 0
 ( 3)  hhsiz_4p = 0

       F(  3,   688) =    0.70
            Prob > F =    0.5516

.  
. ** Column 2: Just cognition index. Limit sample to exclude complexity tretments or consequence message treatment
. reg logspread     cognix_pca                if basesample & !any_complexity & !consequence, robust

Linear regression                               Number of obs     =        708
                                                F(1, 706)         =     111.11
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1570
                                                Root MSE          =     1.9683

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  cognix_pca |  -.8582425    .081419   -10.54   0.000    -1.018095   -.6983902
       _cons |   2.168066   .0730403    29.68   0.000     2.024663    2.311468
------------------------------------------------------------------------------

. 
. ** Column 3: Both. Limit sample to exclude complexity tretments or consequence message treatment
. reg logspread     cognix_pca  $demographics if basesample & !any_complexity & !consequence, robust

Linear regression                               Number of obs     =        708
                                                F(20, 687)        =       7.11
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1879
                                                Root MSE          =     1.9584

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  cognix_pca |  -.8358113   .1053189    -7.94   0.000    -1.042597   -.6290256
         age |   .0316938   .0258373     1.23   0.220    -.0190357    .0824232
       agesq |  -.0349748   .0239159    -1.46   0.144    -.0819318    .0119822
      female |   .0849775   .1638854     0.52   0.604    -.2367989    .4067539
     married |   .0436214   .1796093     0.24   0.808    -.3090276    .3962704
     nhblack |   .1316467   .3716601     0.35   0.723    -.5980793    .8613727
     nhother |   .2930616   .2763656     1.06   0.289    -.2495609    .8356842
    hispanic |  -.5771797   .2752579    -2.10   0.036    -1.117627    -.036732
  ed_dropout |   .7477027   .4250933     1.76   0.079    -.0869353    1.582341
  ed_hschool |   .2752161   .2348559     1.17   0.242    -.1859055    .7363377
  ed_college |  -.0598096   .1946902    -0.31   0.759    -.4420689    .3224497
  ed_graduat |   .3146762   .2161042     1.46   0.146    -.1096279    .7389803
  hinc_25_50 |   .2597673   .2857555     0.91   0.364    -.3012917    .8208263
  hinc_50_75 |   .4856693    .286953     1.69   0.091    -.0777409     1.04908
 hinc_75_100 |  -.1548633   .2999053    -0.52   0.606    -.7437043    .4339776
  hinc_ge100 |  -.0027845   .2549789    -0.01   0.991    -.5034159    .4978469
     hhsiz_2 |  -.1506688   .2305166    -0.65   0.514    -.6032704    .3019329
     hhsiz_3 |   .1407148   .3220607     0.44   0.662    -.4916267    .7730563
    hhsiz_4p |  -.1331193   .3678478    -0.36   0.718    -.8553601    .5891215
     anykids |  -.2043812    .277422    -0.74   0.462    -.7490779    .3403154
       _cons |   1.394785   .7003634     1.99   0.047     .0196757    2.769895
------------------------------------------------------------------------------

. 
. ** Rather than reporting each of these dummies, report p-values on their joint tests of being zero
. testparm nhblack nhother hispanic

 ( 1)  nhblack = 0
 ( 2)  nhother = 0
 ( 3)  hispanic = 0

       F(  3,   687) =    2.13
            Prob > F =    0.0952

. testparm ed_*

 ( 1)  ed_dropout = 0
 ( 2)  ed_hschool = 0
 ( 3)  ed_college = 0
 ( 4)  ed_graduat = 0

       F(  4,   687) =    1.70
            Prob > F =    0.1485

. testparm hinc_*

 ( 1)  hinc_25_50 = 0
 ( 2)  hinc_50_75 = 0
 ( 3)  hinc_75_100 = 0
 ( 4)  hinc_ge100 = 0

       F(  4,   687) =    1.79
            Prob > F =    0.1297

. testparm hhsiz_*

 ( 1)  hhsiz_2 = 0
 ( 2)  hhsiz_3 = 0
 ( 3)  hhsiz_4p = 0

       F(  3,   687) =    0.70
            Prob > F =    0.5524

.                 
.                 
.                 
.                 
.                 
.                 
.         
. // ***************************************************************************************
. //                      APX TABLE A05: MAIN REGRESSIONS, all coefficients reported
. // ***************************************************************************************
.             
. ** APX Table A05 shows the full output of the three regressions of Table 3.
. **
. ** It has the same columns as table , but lots more rows: one row for each explanatory variable
. ** as well as rows for R2 and N.
. 
. ** See the output for Table 3 (reported earlier in the log file)
.     
.         
. 
.         
.         
.         
.         
.         
.         
. // ***************************************************************************************
. //                      APX TABLE A06: MAIN REGRESSION but
. //              Complexity Treatment Split out by Type of Complexity Treatment
. // ***************************************************************************************
.     
. ** This table has the same layout as Table 3 with 2 exceptions:
. **
. ** 1. Instead of a single complexity treatment, there are two rows with different complexity
. **    treatments. These are the first two rows of the table:
. **    Row 1: "Complexity treatment: Wide Spread"  (complexity_2)
. **    Row 2: "Complexity treatment: Added Info"   (complexity_3)
. **
. ** 2. Just above the rows with R2 and N, insert the follow row:
. **    "P-value that coefficients on both complexity treatments are equal"
. 
. 
. ** ------ Column 1 --------
. ** Estimates for rows 1-5, row 7: the p-value on lump-sum shown first (under "P>|t|"), R2 and N are in the output below 
. reg logspread complexity_2 complexity_3 consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =      23.33
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1568
                                                Root MSE          =     1.9605

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   complexity_2 |   .1494204   .0763919     1.96   0.051      -.00035    .2991908
   complexity_3 |   .1139917   .0745133     1.53   0.126    -.0320957     .260079
    consequence |   -.140479   .0616361    -2.28   0.023    -.2613199    -.019638
     cognix_pca |  -.7877918   .0427462   -18.43   0.000    -.8715979   -.7039856
     sell_first |   .1654395    .061665     2.68   0.007      .044542    .2863371
ls_startvalue_2 |   .0636204   .0761229     0.84   0.403    -.0856227    .2128635
ls_startvalue_3 |   -.001235   .0749698    -0.02   0.987    -.1482173    .1457474
       ls_first |   .0302774   .0616314     0.49   0.623    -.0905543     .151109
   ss_benefit_2 |   .1121536   .0870487     1.29   0.198    -.0585101    .2828173
   ss_benefit_3 |   .0567269   .0843089     0.67   0.501    -.1085651    .2220189
   ss_benefit_4 |   .1670934   .0868058     1.92   0.054     -.003094    .3372808
vignette_name_2 |    .114343   .0856407     1.34   0.182    -.0535602    .2822462
vignette_name_3 |    .089812   .0880426     1.02   0.308    -.0828002    .2624241
vignette_name_4 |  -.0098805   .0852609    -0.12   0.908     -.177039     .157278
            age |   .0247629   .0129922     1.91   0.057     -.000709    .0502348
          agesq |  -.0146936   .0129151    -1.14   0.255    -.0400144    .0106272
         female |   .0853478    .066319     1.29   0.198    -.0446742    .2153698
        married |   .0959846   .0759945     1.26   0.207    -.0530068     .244976
        nhblack |    .028033   .1419029     0.20   0.843    -.2501753    .3062412
        nhother |   .0483704   .1215531     0.40   0.691     -.189941    .2866819
       hispanic |    .081131   .1251898     0.65   0.517    -.1643104    .3265723
     ed_dropout |  -.0585211   .1778733    -0.33   0.742    -.4072512     .290209
     ed_hschool |   .0338814   .0930291     0.36   0.716    -.1485072      .21627
     ed_college |   .0071948   .0838438     0.09   0.932    -.1571855    .1715751
     ed_graduat |   .0766582   .0954304     0.80   0.422    -.1104382    .2637546
     hinc_25_50 |   .1020894   .1167399     0.87   0.382    -.1267854    .3309642
     hinc_50_75 |  -.1670493   .1159924    -1.44   0.150    -.3944586      .06036
    hinc_75_100 |  -.0555349   .1300357    -0.43   0.669    -.3104769    .1994071
     hinc_ge100 |  -.2576225   .1099666    -2.34   0.019    -.4732179   -.0420271
        hhsiz_2 |  -.0243049   .0954436    -0.25   0.799    -.2114271    .1628173
        hhsiz_3 |   .1457747    .130605     1.12   0.264    -.1102833    .4018328
       hhsiz_4p |   .1798826   .1453796     1.24   0.216    -.1051418    .4649071
        anykids |  -.1754011   .1056487    -1.66   0.097    -.3825309    .0317288
          _cons |   1.106359   .3422105     3.23   0.001     .4354372    1.777281
---------------------------------------------------------------------------------

.         
. ** Row 6: p-value on starting values
. testparm ls_startvalue_*

 ( 1)  ls_startvalue_2 = 0
 ( 2)  ls_startvalue_3 = 0

       F(  2,  4026) =    0.47
            Prob > F =    0.6238

. 
. ** Row 8: p-value on benefit amounts
. testparm ss_benefit_*

 ( 1)  ss_benefit_2 = 0
 ( 2)  ss_benefit_3 = 0
 ( 3)  ss_benefit_4 = 0

       F(  3,  4026) =    1.38
            Prob > F =    0.2483

. 
. ** Row 9: p-value on benefit amounts
. testparm vignette_name_*

 ( 1)  vignette_name_2 = 0
 ( 2)  vignette_name_3 = 0
 ( 3)  vignette_name_4 = 0

       F(  3,  4026) =    1.04
            Prob > F =    0.3735

. 
. ** Row 11: p-value on complexity treatments being equal
. testparm complexity_*, equal

 ( 1)  - complexity_2 + complexity_3 = 0

       F(  1,  4026) =    0.21
            Prob > F =    0.6461

. 
. 
. 
. ** ------ Column 2 --------
. ** Estimates for rows 1-5, row 7: the p-value on lump-sum shown first (under "P>|t|"), R2 and N are in the output below 
. reg logsellprice complexity_2 complexity_3 consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =       4.46
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0351
                                                Root MSE          =     1.7384

---------------------------------------------------------------------------------
                |               Robust
   logsellprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   complexity_2 |    .066025   .0680009     0.97   0.332    -.0672944    .1993444
   complexity_3 |   .0338436    .065846     0.51   0.607     -.095251    .1629382
    consequence |   .0107029   .0547132     0.20   0.845    -.0965652    .1179711
     cognix_pca |   -.188453   .0381773    -4.94   0.000    -.2633015   -.1136044
     sell_first |  -.0428044   .0544826    -0.79   0.432    -.1496204    .0640116
ls_startvalue_2 |   .2396981   .0666342     3.60   0.000     .1090581     .370338
ls_startvalue_3 |   .4850889   .0676224     7.17   0.000     .3525117    .6176662
       ls_first |  -.0429083   .0548296    -0.78   0.434    -.1504047    .0645881
   ss_benefit_2 |   .0095157   .0754712     0.13   0.900    -.1384496     .157481
   ss_benefit_3 |  -.0065732   .0739529    -0.09   0.929    -.1515619    .1384155
   ss_benefit_4 |  -.1183023   .0798395    -1.48   0.138     -.274832    .0382273
vignette_name_2 |  -.0282405   .0760623    -0.37   0.710    -.1773648    .1208837
vignette_name_3 |  -.0957162   .0765341    -1.25   0.211    -.2457655    .0543331
vignette_name_4 |  -.0801972   .0761847    -1.05   0.293    -.2295614    .0691671
            age |   .0013323   .0108339     0.12   0.902    -.0199082    .0225728
          agesq |   .0057382   .0103622     0.55   0.580    -.0145773    .0260538
         female |  -.0755578   .0584111    -1.29   0.196    -.1900758    .0389602
        married |  -.0074069   .0693909    -0.11   0.915    -.1434515    .1286378
        nhblack |   -.087431   .1343586    -0.65   0.515    -.3508483    .1759863
        nhother |  -.0556313   .1075135    -0.52   0.605    -.2664174    .1551547
       hispanic |  -.0941995   .1215177    -0.78   0.438    -.3324414    .1440425
     ed_dropout |     .13702    .161077     0.85   0.395    -.1787801    .4528201
     ed_hschool |   .1040162   .0854859     1.22   0.224    -.0635835    .2716158
     ed_college |   .0189799   .0730378     0.26   0.795    -.1242146    .1621745
     ed_graduat |   .2238707   .0765977     2.92   0.003     .0736969    .3740446
     hinc_25_50 |   .0464932   .1082769     0.43   0.668    -.1657895    .2587758
     hinc_50_75 |   -.071085    .107484    -0.66   0.508    -.2818132    .1396431
    hinc_75_100 |  -.1047333   .1189842    -0.88   0.379    -.3380082    .1285416
     hinc_ge100 |  -.0990595   .1002679    -0.99   0.323    -.2956402    .0975211
        hhsiz_2 |   .0420941   .0836812     0.50   0.615    -.1219673    .2061554
        hhsiz_3 |   .2514106   .1135202     2.21   0.027     .0288481    .4739731
       hhsiz_4p |   .1764613   .1347842     1.31   0.191    -.0877904    .4407129
        anykids |  -.2374514   .1010309    -2.35   0.019    -.4355279    -.039375
          _cons |   9.345316   .3059717    30.54   0.000     8.745442     9.94519
---------------------------------------------------------------------------------

. 
. 
. ** Row 6: p-value on starting values
. testparm ls_startvalue_*

 ( 1)  ls_startvalue_2 = 0
 ( 2)  ls_startvalue_3 = 0

       F(  2,  4026) =   25.73
            Prob > F =    0.0000

. 
. ** Row 8: p-value on benefit amounts
. testparm ss_benefit_*

 ( 1)  ss_benefit_2 = 0
 ( 2)  ss_benefit_3 = 0
 ( 3)  ss_benefit_4 = 0

       F(  3,  4026) =    1.05
            Prob > F =    0.3679

. 
. ** Row 9: p-value on benefit amounts
. testparm vignette_name_*

 ( 1)  vignette_name_2 = 0
 ( 2)  vignette_name_3 = 0
 ( 3)  vignette_name_4 = 0

       F(  3,  4026) =    0.68
            Prob > F =    0.5662

. 
. ** Row 11: p-value on complexity treatments being equal
. testparm complexity_*, equal

 ( 1)  - complexity_2 + complexity_3 = 0

       F(  1,  4026) =    0.22
            Prob > F =    0.6382

. 
. 
. 
. ** ------ Column 3 --------
. ** Estimates for rows 1-5, row 7: the p-value on lump-sum shown first (under "P>|t|"), R2 and N are in the output below 
. reg logbuyprice complexity_2 complexity_3 consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =       9.81
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0672
                                                Root MSE          =     2.0581

---------------------------------------------------------------------------------
                |               Robust
    logbuyprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   complexity_2 |  -.1169264   .0791422    -1.48   0.140    -.2720889    .0382361
   complexity_3 |  -.1559084   .0787927    -1.98   0.048    -.3103856   -.0014311
    consequence |   .1335382   .0648039     2.06   0.039     .0064866    .2605898
     cognix_pca |   .0980288   .0456476     2.15   0.032     .0085342    .1875234
     sell_first |   .7766235   .0653308    11.89   0.000     .6485391     .904708
ls_startvalue_2 |   .2359888    .079219     2.98   0.003     .0806758    .3913018
ls_startvalue_3 |    .477404   .0788966     6.05   0.000      .322723    .6320849
       ls_first |  -.0640738   .0648058    -0.99   0.323     -.191129    .0629814
   ss_benefit_2 |  -.4591779   .0925657    -4.96   0.000    -.6406579   -.2776978
   ss_benefit_3 |  -.3940901   .0906848    -4.35   0.000    -.5718824   -.2162978
   ss_benefit_4 |  -.3532557   .0929916    -3.80   0.000    -.5355708   -.1709407
vignette_name_2 |  -.0980029   .0886221    -1.11   0.269    -.2717513    .0757455
vignette_name_3 |   .1157596   .0913754     1.27   0.205    -.0633867    .2949059
vignette_name_4 |   .1473991   .0890623     1.66   0.098    -.0272123    .3220105
            age |  -.0346743   .0129377    -2.68   0.007    -.0600393   -.0093093
          agesq |   .0234527   .0128964     1.82   0.069    -.0018313    .0487368
         female |  -.1596782   .0687899    -2.32   0.020    -.2945445   -.0248118
        married |  -.1043633   .0814927    -1.28   0.200    -.2641341    .0554075
        nhblack |  -.1160307   .1478035    -0.79   0.432    -.4058073    .1737459
        nhother |    -.08701    .128376    -0.68   0.498     -.338698     .164678
       hispanic |  -.0967898   .1330049    -0.73   0.467     -.357553    .1639734
     ed_dropout |   .1343958   .1819873     0.74   0.460       -.2224    .4911915
     ed_hschool |   .0481093   .0989373     0.49   0.627    -.1458624    .2420811
     ed_college |   .0995404   .0864868     1.15   0.250    -.0700215    .2691023
     ed_graduat |   .2330997   .0995714     2.34   0.019     .0378847    .4283147
     hinc_25_50 |  -.1356982   .1228653    -1.10   0.269    -.3765821    .1051857
     hinc_50_75 |  -.0383632   .1214901    -0.32   0.752     -.276551    .1998246
    hinc_75_100 |  -.0110363   .1317044    -0.08   0.933    -.2692498    .2471772
     hinc_ge100 |  -.0416392    .111252    -0.37   0.708    -.2597546    .1764761
        hhsiz_2 |  -.0065665   .0996537    -0.07   0.947    -.2019429      .18881
        hhsiz_3 |  -.0262658   .1343122    -0.20   0.845     -.289592    .2370603
       hhsiz_4p |  -.2089286   .1511632    -1.38   0.167    -.5052922     .087435
        anykids |   .1185268    .113576     1.04   0.297    -.1041451    .3411987
          _cons |   9.608313   .3506351    27.40   0.000     8.920874    10.29575
---------------------------------------------------------------------------------

. 
. ** Row 6: p-value on starting values
. testparm ls_startvalue_*

 ( 1)  ls_startvalue_2 = 0
 ( 2)  ls_startvalue_3 = 0

       F(  2,  4026) =   18.31
            Prob > F =    0.0000

. 
. ** Row 8: p-value on benefit amounts
. testparm ss_benefit_*

 ( 1)  ss_benefit_2 = 0
 ( 2)  ss_benefit_3 = 0
 ( 3)  ss_benefit_4 = 0

       F(  3,  4026) =    9.90
            Prob > F =    0.0000

. 
. ** Row 9: p-value on benefit amounts
. testparm vignette_name_*

 ( 1)  vignette_name_2 = 0
 ( 2)  vignette_name_3 = 0
 ( 3)  vignette_name_4 = 0

       F(  3,  4026) =    2.95
            Prob > F =    0.0315

. 
. ** Row 11: p-value on complexity treatments being equal
. testparm complexity_*, equal

 ( 1)  - complexity_2 + complexity_3 = 0

       F(  1,  4026) =    0.24
            Prob > F =    0.6260

. 
. 
. 
. 
. 
. // ***************************************************************************************
. //                      APX TABLE A07: FURTHER HETEROGENEITY IN TREATMENT EFFECTS
. // ***************************************************************************************
. 
. 
. ** ----------- Specification 1: "By Gender" --------------
. **
. ** Two Rows:   1. "Female"
. **             2. "Male"
. **
. 
. assert female<. if basesample

. 
. ** Create interaction variables
. gen anycompXfem1 = any_complexity*(female==1)

. gen anycompXfem0 = any_complexity*(female==0)

. 
. gen consXfem1    = consequence*(female==1)

. gen consXfem0    = consequence*(female==0)

. 
. ** Output for specification 1
. reg logspread   anycompXfem* consXfem* cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(34, 4025)       =      22.64
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1568
                                                Root MSE          =     1.9608

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   anycompXfem1 |   .1258614   .0889905     1.41   0.157    -.0486093    .3003321
   anycompXfem0 |   .1392273   .0932258     1.49   0.135    -.0435469    .3220015
      consXfem1 |  -.1520055   .0855983    -1.78   0.076    -.3198256    .0158145
      consXfem0 |   -.124964   .0880089    -1.42   0.156    -.2975102    .0475822
     cognix_pca |  -.7872511    .042862   -18.37   0.000    -.8712844   -.7032179
     sell_first |   .1659559   .0616851     2.69   0.007     .0450189    .2868929
ls_startvalue_2 |   .0629811   .0761633     0.83   0.408    -.0863412    .2123035
ls_startvalue_3 |  -.0025184   .0749468    -0.03   0.973    -.1494555    .1444187
       ls_first |   .0294902   .0616212     0.48   0.632    -.0913214    .1503019
   ss_benefit_2 |   .1123345   .0870593     1.29   0.197      -.05835     .283019
   ss_benefit_3 |   .0569774   .0842883     0.68   0.499    -.1082743    .2222291
   ss_benefit_4 |   .1668325    .086796     1.92   0.055    -.0033357    .3370008
vignette_name_2 |   .1141013    .085606     1.33   0.183    -.0537339    .2819364
vignette_name_3 |   .0883699   .0878002     1.01   0.314     -.083767    .2605069
vignette_name_4 |  -.0108314   .0851188    -0.13   0.899    -.1777114    .1560485
            age |   .0248511   .0129952     1.91   0.056    -.0006268    .0503289
          agesq |  -.0147922   .0129211    -1.14   0.252    -.0401247    .0105404
         female |   .1081295    .125811     0.86   0.390    -.1385298    .3547887
        married |   .0968956   .0760206     1.27   0.203    -.0521469     .245938
        nhblack |   .0291995   .1419625     0.21   0.837    -.2491256    .3075247
        nhother |   .0486626   .1215686     0.40   0.689    -.1896792    .2870044
       hispanic |   .0816362   .1254325     0.65   0.515    -.1642809    .3275534
     ed_dropout |  -.0578175   .1778678    -0.33   0.745    -.4065369    .2909019
     ed_hschool |   .0332675   .0930321     0.36   0.721     -.149127    .2156619
     ed_college |   .0072922   .0839831     0.09   0.931    -.1573611    .1719455
     ed_graduat |   .0755467   .0955171     0.79   0.429    -.1117197    .2628131
     hinc_25_50 |   .1015215   .1167574     0.87   0.385    -.1273877    .3304307
     hinc_50_75 |  -.1664951   .1159171    -1.44   0.151    -.3937569    .0607666
    hinc_75_100 |  -.0554188   .1300626    -0.43   0.670    -.3104136    .1995759
     hinc_ge100 |  -.2570192   .1099339    -2.34   0.019    -.4725505    -.041488
        hhsiz_2 |  -.0252184   .0956953    -0.26   0.792    -.2128342    .1623974
        hhsiz_3 |   .1461933   .1308379     1.12   0.264    -.1103214    .4027079
       hhsiz_4p |   .1807724   .1454653     1.24   0.214    -.1044201    .4659648
        anykids |  -.1766798   .1056124    -1.67   0.094    -.3837385    .0303789
          _cons |   1.092605   .3474882     3.14   0.002     .4113357    1.773874
---------------------------------------------------------------------------------

.  
. ** P-value in the Complexity treatment column:
. testparm anycompXfem*, equal

 ( 1)  - anycompXfem1 + anycompXfem0 = 0

       F(  1,  4025) =    0.01
            Prob > F =    0.9173

.  
. ** P-value in the Consequence message treatment column: 
. testparm consXfem*, equal

 ( 1)  - consXfem1 + consXfem0 = 0

       F(  1,  4025) =    0.05
            Prob > F =    0.8261

. 
. ** For the nobs for the rows (in square brackets in the last column, check that they add up to the N of the regression)
. tab female if basesample

     Female |      Freq.     Percent        Cum.
------------+-----------------------------------
         No |      1,729       42.59       42.59
        Yes |      2,331       57.41      100.00
------------+-----------------------------------
      Total |      4,060      100.00

. 
. 
. 
. 
. ** --------- Specifcation 2: "By Education" ------------------ 
. **
. ** Two Rows:   1. "Some college or less"
. **             2. "Bachelor's degree or more"
. **
. 
. ** Double check where the median is
. tab edu_ix if basesample

      Education |
     Index, 1-5 |
          Scale |      Freq.     Percent        Cum.
----------------+-----------------------------------
     HS Dropout |        217        5.34        5.34
    High School |        785       19.33       24.68
   Some College |      1,575       38.79       63.47
      Bachelors |        884       21.77       85.25
Graduate Degree |        599       14.75      100.00
----------------+-----------------------------------
          Total |      4,060      100.00

. assert edu_ix<. if basesample

. 
. ** Create interaction variables
. gen anycompXedu0 = any_complexity*(edu_ix<=3)

. gen anycompXedu1 = any_complexity*(edu_ix>=4)

. 
. gen consXedu0    = consequence*(edu_ix<=3)

. gen consXedu1    = consequence*(edu_ix>=4)

. 
. ** Output for specification 2
. reg logspread   anycompXedu* consXedu* cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(34, 4025)       =      22.65
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1569
                                                Root MSE          =     1.9606

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   anycompXedu0 |   .1348808   .0853178     1.58   0.114    -.0323892    .3021508
   anycompXedu1 |   .1223548    .097591     1.25   0.210    -.0689775    .3136871
      consXedu0 |  -.1786275   .0816512    -2.19   0.029    -.3387091    -.018546
      consXedu1 |  -.0743061   .0919552    -0.81   0.419    -.2545892    .1059769
     cognix_pca |   -.788307   .0427322   -18.45   0.000    -.8720857   -.7045283
     sell_first |   .1648849    .061636     2.68   0.008     .0440441    .2857256
ls_startvalue_2 |   .0643323   .0761632     0.84   0.398    -.0849898    .2136544
ls_startvalue_3 |  -.0011405   .0750134    -0.02   0.988    -.1482084    .1459273
       ls_first |    .029156   .0616109     0.47   0.636    -.0916355    .1499475
   ss_benefit_2 |   .1101035   .0870915     1.26   0.206    -.0606439     .280851
   ss_benefit_3 |   .0574271   .0842744     0.68   0.496    -.1077973    .2226515
   ss_benefit_4 |   .1667237   .0868139     1.92   0.055    -.0034797    .3369271
vignette_name_2 |   .1144759   .0856328     1.34   0.181    -.0534119    .2823636
vignette_name_3 |   .0879768   .0880956     1.00   0.318    -.0847393    .2606929
vignette_name_4 |  -.0110883   .0851423    -0.13   0.896    -.1780144    .1558378
            age |   .0247522   .0129828     1.91   0.057    -.0007012    .0502056
          agesq |  -.0146938   .0129053    -1.14   0.255    -.0399952    .0106077
         female |   .0852857   .0663059     1.29   0.198    -.0447105    .2152819
        married |   .0963747   .0760383     1.27   0.205    -.0527025    .2454519
        nhblack |   .0260139   .1417815     0.18   0.854    -.2519564    .3039842
        nhother |   .0500916   .1215536     0.41   0.680    -.1882207    .2884039
       hispanic |   .0828746   .1252162     0.66   0.508    -.1626185    .3283677
     ed_dropout |  -.0584711   .1778587    -0.33   0.742    -.4071726    .2902304
     ed_hschool |   .0322755   .0930292     0.35   0.729    -.1501133    .2146643
     ed_college |  -.0370741   .1337079    -0.28   0.782    -.2992156    .2250673
     ed_graduat |    .033128   .1391851     0.24   0.812    -.2397519    .3060079
     hinc_25_50 |   .1014613   .1167095     0.87   0.385    -.1273539    .3302765
     hinc_50_75 |  -.1685226   .1159149    -1.45   0.146      -.39578    .0587349
    hinc_75_100 |  -.0542018   .1300226    -0.42   0.677    -.3091181    .2007145
     hinc_ge100 |  -.2572238   .1099128    -2.34   0.019    -.4727136    -.041734
        hhsiz_2 |  -.0248875   .0956066    -0.26   0.795    -.2123292    .1625543
        hhsiz_3 |   .1460038   .1305798     1.12   0.264    -.1100049    .4020126
       hhsiz_4p |   .1816176   .1451995     1.25   0.211    -.1030538    .4662891
        anykids |  -.1782257   .1057041    -1.69   0.092    -.3854643    .0290128
          _cons |    1.12663   .3443997     3.27   0.001     .4514159    1.801844
---------------------------------------------------------------------------------

.  
. ** P-value in the Complexity treatment column:
. testparm anycompXedu*, equal

 ( 1)  - anycompXedu0 + anycompXedu1 = 0

       F(  1,  4025) =    0.01
            Prob > F =    0.9231

.  
. ** P-value in the Consequence message treatment column: 
. testparm consXedu*, equal

 ( 1)  - consXedu0 + consXedu1 = 0

       F(  1,  4025) =    0.72
            Prob > F =    0.3970

. 
. ** For the nobs for the rows (in square brackets in the last column, check that they add up to the N of the regression)
. count if  basesample & (edu_ix<=3)
  2,577

. count if  basesample & (edu_ix>=4)
  1,483

. 
. 
. 
. ** ----------- Specification 3: "By Age" ----------------
. **
. ** Two Rows:   1. "Below median (less than 50)"
. **             2. "Above median (50 or more)"
. **
. 
. ** Find the median age 
. sum age if basesample, d

                             Age
-------------------------------------------------------------
      Percentiles      Smallest
 1%           19             18
 5%           25             18
10%           28             18       Obs               4,060
25%           36             18       Sum of Wgt.       4,060

50%           49                      Mean           48.47833
                        Largest       Std. Dev.      15.45153
75%           60            106
90%           69            106       Variance       238.7497
95%           74            106       Skewness       .1559277
99%           82            106       Kurtosis        2.36288

. assert age<. if basesample

. gen age_abovemed = age > r(p50)

. 
. ** Create interaction variables
. gen anycompXage0= any_complexity*(age <= r(p50))

. gen anycompXage1= any_complexity*(age >  r(p50))

. 
. gen consXage0= consequence*(age <= r(p50))

. gen consXage1= consequence*(age >  r(p50))

. 
. ** Output for specification 3
. reg logspread   anycompXage* consXage* age_abovemed cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(35, 4024)       =      22.21
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1577
                                                Root MSE          =       1.96

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   anycompXage0 |   .0215817   .0906645     0.24   0.812     -.156171    .1993344
   anycompXage1 |   .2518746   .0923763     2.73   0.006      .070766    .4329833
      consXage0 |  -.1908056   .0856865    -2.23   0.026    -.3587986   -.0228126
      consXage1 |  -.0827383   .0890714    -0.93   0.353    -.2573676    .0918911
   age_abovemed |  -.1400836   .1612357    -0.87   0.385    -.4561949    .1760278
     cognix_pca |  -.7887664   .0427644   -18.44   0.000    -.8726082   -.7049245
     sell_first |   .1649988    .061678     2.68   0.007     .0440756    .2859219
ls_startvalue_2 |   .0639122   .0761443     0.84   0.401    -.0853728    .2131972
ls_startvalue_3 |   .0014752   .0751066     0.02   0.984    -.1457752    .1487256
       ls_first |   .0313113   .0615786     0.51   0.611    -.0894168    .1520394
   ss_benefit_2 |   .1126435   .0869309     1.30   0.195    -.0577891    .2830761
   ss_benefit_3 |   .0549084   .0842351     0.65   0.515    -.1102389    .2200558
   ss_benefit_4 |   .1670334   .0868427     1.92   0.055    -.0032264    .3372931
vignette_name_2 |   .1165498   .0855686     1.36   0.173     -.051212    .2843117
vignette_name_3 |   .0876219   .0877203     1.00   0.318    -.0843585    .2596024
vignette_name_4 |  -.0081254   .0851092    -0.10   0.924    -.1749866    .1587358
            age |   .0221315   .0138107     1.60   0.109    -.0049452    .0492082
          agesq |  -.0136417   .0129834    -1.05   0.293    -.0390963     .011813
         female |   .0843159   .0662943     1.27   0.204    -.0456576    .2142894
        married |   .0983327   .0759194     1.30   0.195    -.0505113    .2471767
        nhblack |   .0352147   .1417955     0.25   0.804     -.242783    .3132124
        nhother |   .0498229    .121594     0.41   0.682    -.1885687    .2882144
       hispanic |   .0784535   .1256624     0.62   0.532    -.1679144    .3248214
     ed_dropout |   -.054947   .1776575    -0.31   0.757     -.403254      .29336
     ed_hschool |   .0323628   .0931323     0.35   0.728    -.1502281    .2149537
     ed_college |    .015056   .0837814     0.18   0.857    -.1492019     .179314
     ed_graduat |   .0800241   .0954718     0.84   0.402    -.1071535    .2672018
     hinc_25_50 |   .1007179   .1167631     0.86   0.388    -.1282024    .3296383
     hinc_50_75 |  -.1591839   .1159977    -1.37   0.170    -.3866035    .0682358
    hinc_75_100 |  -.0522834    .129926    -0.40   0.687    -.3070103    .2024435
     hinc_ge100 |  -.2538286   .1100196    -2.31   0.021     -.469528   -.0381292
        hhsiz_2 |  -.0262163   .0953299    -0.28   0.783    -.2131158    .1606831
        hhsiz_3 |   .1448534   .1306785     1.11   0.268    -.1113489    .4010556
       hhsiz_4p |   .1783665   .1451728     1.23   0.219    -.1062527    .4629856
        anykids |   -.162641   .1077875    -1.51   0.131    -.3739642    .0486821
          _cons |   1.261903    .366239     3.45   0.001     .5438716    1.979934
---------------------------------------------------------------------------------

.  
. ** P-value in the Complexity treatment column:
. testparm anycompXage*, equal

 ( 1)  - anycompXage0 + anycompXage1 = 0

       F(  1,  4024) =    3.18
            Prob > F =    0.0746

.  
. ** P-value in the Consequence message treatment column: 
. testparm consXage*, equal

 ( 1)  - consXage0 + consXage1 = 0

       F(  1,  4024) =    0.76
            Prob > F =    0.3828

. 
. ** For the nobs for the rows (in square brackets in the last column, check that they add up to the N of the regression)
. tab age_abovemed if basesample

age_aboveme |
          d |      Freq.     Percent        Cum.
------------+-----------------------------------
          0 |      2,107       51.90       51.90
          1 |      1,953       48.10      100.00
------------+-----------------------------------
      Total |      4,060      100.00

. 
. 
. ** ------------- Specification 4: "By Income" --------------
. **
. ** Two Rows:   1. "Below median (less than $75k)"
. **             2. "Above median ($75k or more)"
. **
. 
. ** Double check where the median is
. sum income_cat if basesample, d

                   Income categories (1-5)
-------------------------------------------------------------
      Percentiles      Smallest
 1%            1              1
 5%            1              1
10%            1              1       Obs               4,060
25%            2              1       Sum of Wgt.       4,060

50%            3                      Mean           3.351478
                        Largest       Std. Dev.      1.517204
75%            5              5
90%            5              5       Variance       2.301907
95%            5              5       Skewness      -.2652238
99%            5              5       Kurtosis       1.579113

. assert income_cat<. if basesample

. gen inc_abovemed = income_cat > 3

. 
. ** Create interaction variables
. gen anycompXinc0= any_complexity*(1-inc_abovemed)

. gen anycompXinc1= any_complexity*(inc_abovemed)

. 
. gen consXinc0= consequence*(1-inc_abovemed)

. gen consXinc1= consequence*(inc_abovemed)

. 
. ** Output for specification 4
. reg logspread   anycompXinc* consXinc* inc_abovemed cognix_pca $exp_controls $demographics if basesample, robust
note: hinc_ge100 omitted because of collinearity

Linear regression                               Number of obs     =      4,060
                                                F(34, 4025)       =      22.59
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1573
                                                Root MSE          =     1.9602

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   anycompXinc0 |   .0738373   .0974077     0.76   0.448    -.1171358    .2648104
   anycompXinc1 |   .1859473   .0855063     2.17   0.030     .0183076     .353587
      consXinc0 |  -.2203982   .0914733    -2.41   0.016    -.3997365   -.0410599
      consXinc1 |  -.0603409    .082952    -0.73   0.467    -.2229727    .1022908
   inc_abovemed |  -.4133047   .1527383    -2.71   0.007    -.7127563    -.113853
     cognix_pca |  -.7871653   .0427668   -18.41   0.000    -.8710118   -.7033187
     sell_first |   .1646916   .0616701     2.67   0.008      .043784    .2855992
ls_startvalue_2 |   .0654351   .0762137     0.86   0.391    -.0839859    .2148561
ls_startvalue_3 |   -.003204   .0749995    -0.04   0.966    -.1502445    .1438366
       ls_first |   .0294466   .0616027     0.48   0.633    -.0913287     .150222
   ss_benefit_2 |   .1142276   .0870199     1.31   0.189    -.0563795    .2848347
   ss_benefit_3 |   .0615687   .0842481     0.73   0.465    -.1036042    .2267416
   ss_benefit_4 |      .1705    .086892     1.96   0.050     .0001435    .3408566
vignette_name_2 |   .1114525   .0856058     1.30   0.193    -.0563822    .2792872
vignette_name_3 |   .0867816     .08779     0.99   0.323    -.0853354    .2588986
vignette_name_4 |  -.0113329   .0850916    -0.13   0.894    -.1781597    .1554938
            age |   .0246512   .0129687     1.90   0.057    -.0007747     .050077
          agesq |  -.0145943   .0128898    -1.13   0.258    -.0398654    .0106768
         female |   .0847413   .0663133     1.28   0.201    -.0452695     .214752
        married |   .0975178    .076161     1.28   0.200    -.0517999    .2468355
        nhblack |   .0313444   .1418717     0.22   0.825    -.2468027    .3094914
        nhother |   .0475549   .1215218     0.39   0.696    -.1906951     .285805
       hispanic |   .0829488   .1252454     0.66   0.508    -.1626016    .3284991
     ed_dropout |  -.0576506   .1776514    -0.32   0.746    -.4059456    .2906444
     ed_hschool |   .0384883   .0930681     0.41   0.679    -.1439767    .2209534
     ed_college |   .0087307   .0838763     0.10   0.917    -.1557133    .1731748
     ed_graduat |   .0807329   .0955264     0.85   0.398    -.1065516    .2680175
     hinc_25_50 |   .0986523   .1167028     0.85   0.398    -.1301498    .3274545
     hinc_50_75 |  -.1669844   .1159214    -1.44   0.150    -.3942546    .0602857
    hinc_75_100 |   .2010738   .1017746     1.98   0.048     .0015393    .4006084
     hinc_ge100 |          0  (omitted)
        hhsiz_2 |  -.0194624    .095492    -0.20   0.839    -.2066795    .1677548
        hhsiz_3 |   .1523388   .1306908     1.17   0.244    -.1038874     .408565
       hhsiz_4p |   .1864028   .1452851     1.28   0.200    -.0984364    .4712421
        anykids |  -.1782135   .1056636    -1.69   0.092    -.3853726    .0289456
          _cons |   1.183208   .3455299     3.42   0.001     .5057786    1.860638
---------------------------------------------------------------------------------

.  
. ** P-value in the Complexity treatment column:
. testparm anycompXinc*, equal

 ( 1)  - anycompXinc0 + anycompXinc1 = 0

       F(  1,  4025) =    0.75
            Prob > F =    0.3866

.  
. ** P-value in the Consequence message treatment column: 
. testparm consXinc*, equal

 ( 1)  - consXinc0 + consXinc1 = 0

       F(  1,  4025) =    1.67
            Prob > F =    0.1963

. 
. ** For the nobs for the rows (in square brackets in the last column, check that they add up to the N of the regression)
. tab inc_abovemed if basesample

inc_aboveme |
          d |      Freq.     Percent        Cum.
------------+-----------------------------------
          0 |      2,054       50.59       50.59
          1 |      2,006       49.41      100.00
------------+-----------------------------------
      Total |      4,060      100.00

. 
. 
. 
. 
. 
.   
. // ***************************************************************************************
. //                      APX TABLE A08: Robustness of the Main Treatment Effects
. // ***************************************************************************************
.         
.         
. ** Row 1: "Baseline" (the same regression as column 1 of Table 3)
. reg logspread    any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      24.04
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1568
                                                Root MSE          =     1.9603

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1314293   .0648348     2.03   0.043     .0043173    .2585414
    consequence |  -.1405306   .0616294    -2.28   0.023    -.2613583   -.0197029
     cognix_pca |  -.7876547   .0427562   -18.42   0.000    -.8714804   -.7038289
     sell_first |   .1657743   .0616662     2.69   0.007     .0448745    .2866741
ls_startvalue_2 |   .0632191   .0761058     0.83   0.406    -.0859904    .2124286
ls_startvalue_3 |  -.0021141   .0749548    -0.03   0.978    -.1490669    .1448388
       ls_first |   .0294388   .0616149     0.48   0.633    -.0913606    .1502381
   ss_benefit_2 |    .112784   .0870307     1.30   0.195    -.0578444    .2834124
   ss_benefit_3 |   .0573012   .0842277     0.68   0.496    -.1078317     .222434
   ss_benefit_4 |   .1669049   .0868148     1.92   0.055    -.0033001    .3371099
vignette_name_2 |   .1140634      .0856     1.33   0.183    -.0537599    .2818867
vignette_name_3 |   .0882214   .0877756     1.01   0.315    -.0838674    .2603102
vignette_name_4 |  -.0109552   .0850845    -0.13   0.898    -.1777678    .1558574
            age |   .0247686   .0129824     1.91   0.056    -.0006842    .0502213
          agesq |  -.0147106   .0129063    -1.14   0.254     -.040014    .0105928
         female |   .0854357   .0663126     1.29   0.198    -.0445738    .2154451
        married |   .0965551   .0759943     1.27   0.204    -.0524358    .2455459
        nhblack |   .0281964   .1418699     0.20   0.842    -.2499472      .30634
        nhother |   .0481969   .1215336     0.40   0.692    -.1900762      .28647
       hispanic |   .0810356   .1251889     0.65   0.517    -.1644039    .3264752
     ed_dropout |  -.0573967   .1777837    -0.32   0.747     -.405951    .2911577
     ed_hschool |   .0334476   .0930188     0.36   0.719    -.1489207    .2158159
     ed_college |   .0076616   .0838392     0.09   0.927    -.1567096    .1720329
     ed_graduat |   .0763065   .0954393     0.80   0.424    -.1108073    .2634204
     hinc_25_50 |   .1019141   .1167251     0.87   0.383    -.1269317      .33076
     hinc_50_75 |  -.1660551    .115919    -1.43   0.152    -.3933204    .0612103
    hinc_75_100 |  -.0550164   .1299977    -0.42   0.672    -.3098838     .199851
     hinc_ge100 |  -.2568218   .1099252    -2.34   0.020    -.4723359   -.0413077
        hhsiz_2 |  -.0247458   .0954113    -0.26   0.795    -.2118047    .1623132
        hhsiz_3 |   .1468662   .1305622     1.12   0.261    -.1091079    .4028402
       hhsiz_4p |   .1815883   .1452072     1.25   0.211    -.1030981    .4662747
        anykids |  -.1769094   .1055735    -1.68   0.094    -.3838918    .0300729
          _cons |   1.106907   .3419853     3.24   0.001     .4364261    1.777387
---------------------------------------------------------------------------------

. 
. 
. 
. ** Panel A: Changing Cognition Measures
. ** ------------------------------------
. 
. ** Row 2: "Cognition score is simple average"
. reg logspread    any_complexity consequence cognix_avg $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      23.73
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1554
                                                Root MSE          =     1.9619

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1309489   .0648858     2.02   0.044     .0037369     .258161
    consequence |  -.1407004   .0616818    -2.28   0.023    -.2616308   -.0197701
     cognix_avg |  -.7808548   .0429269   -18.19   0.000    -.8650152   -.6966943
     sell_first |   .1650349   .0617179     2.67   0.008     .0440336    .2860362
ls_startvalue_2 |   .0636882   .0761585     0.84   0.403    -.0856246     .213001
ls_startvalue_3 |  -.0016523   .0750104    -0.02   0.982    -.1487141    .1454096
       ls_first |   .0299505   .0616691     0.49   0.627    -.0909551    .1508561
   ss_benefit_2 |   .1132933   .0870781     1.30   0.193     -.057428    .2840145
   ss_benefit_3 |   .0571816    .084286     0.68   0.498    -.1080656    .2224287
   ss_benefit_4 |   .1662766   .0868826     1.91   0.056    -.0040614    .3366147
vignette_name_2 |   .1155028   .0856381     1.35   0.177    -.0523952    .2834008
vignette_name_3 |   .0887975   .0878446     1.01   0.312    -.0834265    .2610216
vignette_name_4 |  -.0111696   .0851681    -0.13   0.896    -.1781463     .155807
            age |   .0259417   .0130155     1.99   0.046      .000424    .0514593
          agesq |  -.0154401   .0129384    -1.19   0.233    -.0408065    .0099263
         female |   .0933741   .0663016     1.41   0.159    -.0366138     .223362
        married |   .0960354   .0760674     1.26   0.207    -.0530988    .2451696
        nhblack |   .0260529   .1419474     0.18   0.854    -.2522425    .3043483
        nhother |   .0432653   .1216861     0.36   0.722    -.1953068    .2818375
       hispanic |   .0815926   .1253635     0.65   0.515    -.1641892    .3273743
     ed_dropout |  -.0556248   .1779085    -0.31   0.755    -.4044239    .2931742
     ed_hschool |   .0353586   .0931142     0.38   0.704    -.1471968     .217914
     ed_college |  -.0022813   .0838283    -0.03   0.978    -.1666312    .1620685
     ed_graduat |   .0641204   .0954346     0.67   0.502    -.1229841     .251225
     hinc_25_50 |   .1016892   .1168389     0.87   0.384    -.1273797     .330758
     hinc_50_75 |  -.1678989   .1159948    -1.45   0.148    -.3953128    .0595151
    hinc_75_100 |  -.0578923   .1300449    -0.45   0.656    -.3128523    .1970676
     hinc_ge100 |  -.2601907   .1100123    -2.37   0.018    -.4758758   -.0445057
        hhsiz_2 |  -.0244983   .0955092    -0.26   0.798     -.211749    .1627525
        hhsiz_3 |   .1462785   .1306385     1.12   0.263    -.1098452    .4024022
       hhsiz_4p |   .1818893   .1453042     1.25   0.211    -.1029873    .4667659
        anykids |  -.1783845    .105621    -1.69   0.091      -.38546    .0286911
          _cons |   1.070874   .3428568     3.12   0.002     .3986847    1.743063
---------------------------------------------------------------------------------

. 
. ** Row 3: "All five components of cognition score entered separately"
. reg logspread    any_complexity consequence cog_fin cog_n1 cog_n2 cog_v1 cog_v2 $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(36, 4023)       =      22.82
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1614
                                                Root MSE          =     1.9559

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1314408   .0647104     2.03   0.042     .0045726    .2583091
    consequence |  -.1373987   .0615593    -2.23   0.026    -.2580889   -.0167084
        cog_fin |  -.2183347    .047968    -4.55   0.000    -.3123784   -.1242909
         cog_n1 |  -.3742577   .0476219    -7.86   0.000     -.467623   -.2808924
         cog_n2 |  -.1758886   .0426368    -4.13   0.000    -.2594803    -.092297
         cog_v1 |    -.03888   .0425811    -0.91   0.361    -.1223626    .0446026
         cog_v2 |   -.189779   .0409407    -4.64   0.000    -.2700455   -.1095125
     sell_first |   .1670892   .0615254     2.72   0.007     .0464653    .2877132
ls_startvalue_2 |    .059128   .0761291     0.78   0.437    -.0901272    .2083832
ls_startvalue_3 |  -.0053777   .0747971    -0.07   0.943    -.1520215     .141266
       ls_first |   .0267439   .0614936     0.43   0.664    -.0938177    .1473055
   ss_benefit_2 |   .1100279   .0869654     1.27   0.206    -.0604725    .2805283
   ss_benefit_3 |   .0615623    .084164     0.73   0.465    -.1034457    .2265703
   ss_benefit_4 |    .172042   .0866498     1.99   0.047     .0021604    .3419236
vignette_name_2 |   .1043587   .0857062     1.22   0.223     -.063673    .2723903
vignette_name_3 |   .0860246   .0876387     0.98   0.326    -.0857958    .2578449
vignette_name_4 |  -.0076534     .08483    -0.09   0.928    -.1739672    .1586605
            age |   .0164285   .0131804     1.25   0.213    -.0094125    .0422694
          agesq |  -.0096486   .0129021    -0.75   0.455    -.0349439    .0156466
         female |   .0553797   .0676893     0.82   0.413    -.0773289    .1880883
        married |    .098054   .0758494     1.29   0.196    -.0506528    .2467608
        nhblack |   .0776259   .1426327     0.54   0.586    -.2020132    .3572651
        nhother |   .0879908   .1212033     0.73   0.468    -.1496347    .3256164
       hispanic |    .101776   .1251987     0.81   0.416    -.1436828    .3472349
     ed_dropout |  -.0349578   .1780541    -0.20   0.844    -.3840425    .3141269
     ed_hschool |    .039096   .0931458     0.42   0.675    -.1435213    .2217134
     ed_college |   .0428051   .0845387     0.51   0.613    -.1229376    .2085478
     ed_graduat |   .1207004    .095651     1.26   0.207    -.0668286    .3082293
     hinc_25_50 |   .0927352   .1164686     0.80   0.426    -.1356077    .3210782
     hinc_50_75 |  -.1695359   .1157914    -1.46   0.143    -.3965512    .0574794
    hinc_75_100 |  -.0544687   .1301185    -0.42   0.676     -.309573    .2006355
     hinc_ge100 |  -.2540121   .1102223    -2.30   0.021    -.4701089   -.0379153
        hhsiz_2 |  -.0335695   .0953813    -0.35   0.725    -.2205696    .1534306
        hhsiz_3 |   .1409682    .130547     1.08   0.280    -.1149762    .3969127
       hhsiz_4p |   .1683452   .1453574     1.16   0.247    -.1166358    .4533262
        anykids |  -.1662313   .1058968    -1.57   0.117    -.3738476    .0413849
          _cons |   1.381361    .351026     3.94   0.000     .6931558    2.069566
---------------------------------------------------------------------------------

. 
. ** Row 4: "Financial literacy is only cognition measure"
. reg logspread    any_complexity consequence cog_fin                             $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      16.25
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1146
                                                Root MSE          =     2.0087

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1074607   .0664965     1.62   0.106    -.0229093    .2378307
    consequence |   -.128236   .0632733    -2.03   0.043    -.2522866   -.0041854
        cog_fin |   -.498085   .0439533   -11.33   0.000    -.5842578   -.4119121
     sell_first |   .1313459   .0632596     2.08   0.038     .0073222    .2553696
ls_startvalue_2 |   .0598085   .0778733     0.77   0.443    -.0928662    .2124831
ls_startvalue_3 |  -.0054202   .0770052    -0.07   0.944    -.1563931    .1455526
       ls_first |   .0539193   .0631385     0.85   0.393    -.0698671    .1777057
   ss_benefit_2 |   .1129703   .0889839     1.27   0.204    -.0614873     .287428
   ss_benefit_3 |   .0660055   .0863031     0.76   0.444    -.1031963    .2352073
   ss_benefit_4 |   .1671787   .0892352     1.87   0.061    -.0077717    .3421291
vignette_name_2 |   .0648264   .0872387     0.74   0.457    -.1062098    .2358626
vignette_name_3 |   .0453722   .0900707     0.50   0.614    -.1312162    .2219606
vignette_name_4 |  -.0318257   .0871968    -0.36   0.715    -.2027797    .1391283
            age |   .0222862   .0133515     1.67   0.095    -.0038902    .0484625
          agesq |  -.0095518   .0132478    -0.72   0.471    -.0355249    .0164212
         female |   .2161795   .0676532     3.20   0.001     .0835419    .3488171
        married |   .1216443   .0777611     1.56   0.118    -.0308105    .2740992
        nhblack |   .5192256   .1396665     3.72   0.000     .2454021    .7930492
        nhother |   .1923166   .1260403     1.53   0.127    -.0547921    .4394253
       hispanic |   .3702875   .1263456     2.93   0.003     .1225802    .6179947
     ed_dropout |   .1723971   .1826633     0.94   0.345     -.185724    .5305183
     ed_hschool |   .1421326    .094173     1.51   0.131    -.0424986    .3267638
     ed_college |  -.1610002   .0854276    -1.88   0.060    -.3284855    .0064851
     ed_graduat |  -.1718365    .097644    -1.76   0.079    -.3632728    .0195998
     hinc_25_50 |   .0475966   .1188633     0.40   0.689    -.1854412    .2806344
     hinc_50_75 |  -.2736479   .1178263    -2.32   0.020    -.5046525   -.0426433
    hinc_75_100 |  -.1251676   .1326392    -0.94   0.345    -.3852138    .1348786
     hinc_ge100 |  -.3349722   .1123035    -2.98   0.003    -.5551493   -.1147952
        hhsiz_2 |  -.1129773    .097155    -1.16   0.245    -.3034549    .0775003
        hhsiz_3 |   .0532282   .1332984     0.40   0.690    -.2081105    .3145668
       hhsiz_4p |   .0427701   .1477638     0.29   0.772    -.2469287    .3324689
        anykids |  -.1608636   .1077704    -1.49   0.136    -.3721531    .0504259
          _cons |   1.155446   .3531484     3.27   0.001     .4630799    1.847812
---------------------------------------------------------------------------------

. 
. ** Row 5: "Numeracy measures are only cognition measures"
. reg logspread    any_complexity consequence         cog_n1 cog_n2               $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =      23.63
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1495
                                                Root MSE          =      1.969

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1207085   .0650343     1.86   0.064    -.0067947    .2482117
    consequence |  -.1526724   .0618808    -2.47   0.014     -.273993   -.0313518
         cog_n1 |  -.4967747   .0438377   -11.33   0.000    -.5827209   -.4108285
         cog_n2 |  -.2555551   .0419012    -6.10   0.000    -.3377046   -.1734056
     sell_first |   .1673493   .0619362     2.70   0.007       .04592    .2887786
ls_startvalue_2 |   .0471079   .0764891     0.62   0.538     -.102853    .1970687
ls_startvalue_3 |   -.033598    .075414    -0.45   0.656    -.1814511    .1142551
       ls_first |   .0339345   .0618545     0.55   0.583    -.0873346    .1552036
   ss_benefit_2 |    .127283   .0875853     1.45   0.146    -.0444326    .2989986
   ss_benefit_3 |   .0622589   .0847105     0.73   0.462    -.1038206    .2283385
   ss_benefit_4 |   .1761176   .0874978     2.01   0.044     .0045735    .3476617
vignette_name_2 |   .1125127   .0862646     1.30   0.192    -.0566136     .281639
vignette_name_3 |   .0750418   .0877852     0.85   0.393    -.0970658    .2471495
vignette_name_4 |   -.010893   .0857715    -0.13   0.899    -.1790526    .1572666
            age |   .0050535    .012856     0.39   0.694    -.0201513    .0302583
          agesq |  -.0013784   .0127636    -0.11   0.914    -.0264022    .0236454
         female |   .0822517    .067124     1.23   0.221    -.0493485    .2138518
        married |   .0880366   .0759349     1.16   0.246    -.0608378    .2369109
        nhblack |   .2534189   .1405718     1.80   0.071    -.0221796    .5290174
        nhother |   .1672536   .1211242     1.38   0.167    -.0702168    .4047241
       hispanic |    .186822    .125382     1.49   0.136    -.0589961    .4326402
     ed_dropout |   .1387407   .1750645     0.79   0.428    -.2044826     .481964
     ed_hschool |   .1394772    .092353     1.51   0.131    -.0415857    .3205402
     ed_college |  -.0308756   .0840551    -0.37   0.713    -.1956702     .133919
     ed_graduat |    .038477   .0958881     0.40   0.688    -.1495167    .2264708
     hinc_25_50 |   .0462366   .1169368     0.40   0.693    -.1830243    .2754975
     hinc_50_75 |  -.2211142   .1156549    -1.91   0.056    -.4478617    .0056334
    hinc_75_100 |  -.1129412   .1305411    -0.87   0.387     -.368874    .1429917
     hinc_ge100 |  -.3446721   .1095878    -3.15   0.002    -.5595248   -.1298193
        hhsiz_2 |  -.0624464   .0951885    -0.66   0.512    -.2490684    .1241757
        hhsiz_3 |    .127481   .1310393     0.97   0.331    -.1294286    .3843905
       hhsiz_4p |   .1467628   .1454475     1.01   0.313    -.1383948    .4319203
        anykids |  -.1605531   .1063942    -1.51   0.131    -.3691445    .0480384
          _cons |   1.773981   .3395486     5.22   0.000     1.108278    2.439685
---------------------------------------------------------------------------------

. 
. ** Row 6: "Verbal measures are only cognition measures"
. reg logspread    any_complexity consequence                       cog_v1 cog_v2 $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =      16.75
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1174
                                                Root MSE          =     2.0058

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |    .111381   .0663682     1.68   0.093    -.0187375    .2414994
    consequence |  -.1522224   .0630577    -2.41   0.016    -.2758504   -.0285944
         cog_v1 |  -.1521621    .042728    -3.56   0.000    -.2359326   -.0683917
         cog_v2 |  -.3830815   .0394788    -9.70   0.000    -.4604818   -.3056812
     sell_first |   .1492794   .0630764     2.37   0.018     .0256148     .272944
ls_startvalue_2 |   .0458639   .0776863     0.59   0.555    -.1064443    .1981721
ls_startvalue_3 |  -.0168574    .076683    -0.22   0.826    -.1671985    .1334837
       ls_first |   .0349265   .0631086     0.55   0.580    -.0888012    .1586543
   ss_benefit_2 |   .1269069   .0887968     1.43   0.153    -.0471839    .3009978
   ss_benefit_3 |   .0677065    .086045     0.79   0.431    -.1009894    .2364024
   ss_benefit_4 |   .1625549   .0887282     1.83   0.067    -.0114014    .3365113
vignette_name_2 |    .128398   .0872708     1.47   0.141     -.042701     .299497
vignette_name_3 |   .0967124   .0896314     1.08   0.281    -.0790147    .2724395
vignette_name_4 |   -.012142   .0874846    -0.14   0.890    -.1836602    .1593762
            age |   .0224131   .0137232     1.63   0.102    -.0044919    .0493181
          agesq |  -.0149412   .0135729    -1.10   0.271    -.0415517    .0116693
         female |   .3272081   .0662507     4.94   0.000       .19732    .4570962
        married |   .0763197   .0779446     0.98   0.328    -.0764948    .2291341
        nhblack |   .2763279   .1417556     1.95   0.051    -.0015916    .5542475
        nhother |   .1051614   .1244974     0.84   0.398    -.1389224    .3492453
       hispanic |    .262413   .1274497     2.06   0.040     .0125411    .5122849
     ed_dropout |   .2079474    .176407     1.18   0.239    -.1379079    .5538028
     ed_hschool |   .1913456   .0943575     2.03   0.043     .0063528    .3763385
     ed_college |  -.2412854   .0838604    -2.88   0.004    -.4056981   -.0768727
     ed_graduat |  -.2478352   .0950351    -2.61   0.009    -.4341566   -.0615138
     hinc_25_50 |   .0442556   .1196368     0.37   0.711    -.1902986    .2788099
     hinc_50_75 |  -.2898655    .117828    -2.46   0.014    -.5208737   -.0588574
    hinc_75_100 |  -.2013757   .1319342    -1.53   0.127    -.4600397    .0572882
     hinc_ge100 |  -.4442169   .1119483    -3.97   0.000    -.6636976   -.2247362
        hhsiz_2 |  -.0445572   .0983184    -0.45   0.650    -.2373156    .1482012
        hhsiz_3 |   .0967492   .1334785     0.72   0.469    -.1649425    .3584409
       hhsiz_4p |   .1334741   .1482785     0.90   0.368    -.1572338     .424182
        anykids |  -.1576439   .1078249    -1.46   0.144    -.3690404    .0537526
          _cons |   1.282586   .3619259     3.54   0.000     .5730111    1.992161
---------------------------------------------------------------------------------

. 
. ** Row 7: Additional controls for cognition, knowledge, and financial experience 
. reg logspread    any_complexity consequence ///
>                  admc__Sframe* admc__Stime* admc__Ssubset* ///
>                  ssa__Sknowledge* ssa__Sliteracy* ssa__Sconfident* ///
>                  afin__annuity* afin__irakeogh* ///
>                  aplan__index* ///
>                  cognix_pca $exp_controls $demographics if basesample, robust
note: admc__Stime_m omitted because of collinearity
note: admc__Ssubset_m omitted because of collinearity

Linear regression                               Number of obs     =      4,060
                                                F(48, 4011)       =      16.46
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1601
                                                Root MSE          =     1.9603

-----------------------------------------------------------------------------------
                  |               Robust
        logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
------------------+----------------------------------------------------------------
   any_complexity |   .1347386   .0648823     2.08   0.038     .0075332     .261944
      consequence |  -.1357264   .0618035    -2.20   0.028    -.2568956   -.0145573
   admc__Sframe_m |   1.007624   1.236432     0.81   0.415    -1.416471    3.431718
     admc__Sframe |  -.0569727   .0358987    -1.59   0.113     -.127354    .0134087
    admc__Stime_m |          0  (omitted)
      admc__Stime |   .0233266   .0334775     0.70   0.486    -.0423079    .0889611
  admc__Ssubset_m |          0  (omitted)
    admc__Ssubset |   .0021313   .0357078     0.06   0.952    -.0678759    .0721384
ssa__Sknowledge_m |    -.54951    .341186    -1.61   0.107    -1.218424    .1194041
  ssa__Sknowledge |   .0072749   .0378346     0.19   0.848    -.0669019    .0814518
 ssa__Sliteracy_m |   .5924611   .3528077     1.68   0.093    -.0992379     1.28416
   ssa__Sliteracy |  -.0560073   .0384254    -1.46   0.145    -.1313425    .0193278
ssa__Sconfident_m |  -.1325974   .1030306    -1.29   0.198    -.3345946    .0693999
  ssa__Sconfident |   -.052185   .0385194    -1.35   0.176    -.1277044    .0233345
  afin__annuity_m |   .3480293   .8738467     0.40   0.690    -1.365196    2.061254
    afin__annuity |   .0084003   .2175828     0.04   0.969    -.4181829    .4349835
 afin__irakeogh_m |  -.2924786   .8798419    -0.33   0.740    -2.017458      1.4325
   afin__irakeogh |      .0211   .0729605     0.29   0.772    -.1219431    .1641431
   aplan__index_m |  -.0302268   .1079535    -0.28   0.779    -.2418756    .1814221
     aplan__index |   -.019606   .0420108    -0.47   0.641    -.1019706    .0627585
       cognix_pca |  -.7590652   .0462893   -16.40   0.000    -.8498179   -.6683125
       sell_first |   .1653396   .0617409     2.68   0.007     .0442932    .2863861
  ls_startvalue_2 |   .0605993   .0762537     0.79   0.427    -.0889003    .2100988
  ls_startvalue_3 |  -.0008038   .0752494    -0.01   0.991    -.1483344    .1467269
         ls_first |   .0317619   .0615867     0.52   0.606    -.0889823     .152506
     ss_benefit_2 |   .1105945   .0871814     1.27   0.205    -.0603295    .2815185
     ss_benefit_3 |   .0535008   .0845169     0.63   0.527    -.1121992    .2192009
     ss_benefit_4 |   .1557542   .0870975     1.79   0.074    -.0150054    .3265137
  vignette_name_2 |   .1095426    .085661     1.28   0.201    -.0584005    .2774858
  vignette_name_3 |    .085461   .0879263     0.97   0.331    -.0869235    .2578455
  vignette_name_4 |  -.0176146   .0851578    -0.21   0.836    -.1845712    .1493419
              age |   .0233674   .0138116     1.69   0.091    -.0037109    .0504458
            agesq |  -.0090995   .0142926    -0.64   0.524    -.0371209    .0189219
           female |   .0818862   .0671674     1.22   0.223    -.0497992    .2135717
          married |   .0904347   .0769713     1.17   0.240    -.0604718    .2413413
          nhblack |   .0373407   .1427345     0.26   0.794    -.2424983    .3171797
          nhother |   .0455899   .1215226     0.38   0.708    -.1926618    .2838417
         hispanic |   .0934616    .129937     0.72   0.472    -.1612872    .3482104
       ed_dropout |   -.045773    .178803    -0.26   0.798    -.3963263    .3047803
       ed_hschool |   .0387037   .0931199     0.42   0.678     -.143863    .2212705
       ed_college |   .0106532    .085483     0.12   0.901    -.1569409    .1782473
       ed_graduat |   .0853051   .0967628     0.88   0.378    -.1044036    .2750139
       hinc_25_50 |   .0884428    .117115     0.76   0.450    -.1411677    .3180534
       hinc_50_75 |   -.193941   .1168594    -1.66   0.097    -.4230504    .0351685
      hinc_75_100 |  -.0811081   .1314982    -0.62   0.537    -.3389176    .1767014
       hinc_ge100 |  -.2766417   .1126993    -2.45   0.014    -.4975949   -.0556885
          hhsiz_2 |  -.0234231   .0958213    -0.24   0.807    -.2112862      .16444
          hhsiz_3 |    .129884   .1305935     0.99   0.320    -.1261519    .3859199
         hhsiz_4p |   .1557493   .1444033     1.08   0.281    -.1273614    .4388601
          anykids |  -.1586431   .1051362    -1.51   0.131    -.3647685    .0474822
            _cons |   1.089681   .3618927     3.01   0.003     .3801701    1.799192
-----------------------------------------------------------------------------------

. 
. 
. 
. 
. ** Panel B: Sample Selection
. ** ------------------------------------
. 
. ** Reminder: gen byte basesample = ~(miss_spread | miss_anydemographic | miss_cognix_pca)                       
. 
. ** Dummy out the demographics
. foreach var of varlist $demographics {
  2.     gen DQ_`var'=`var'
  3.     gen byte DM_`var' = missing(`var')
  4.     replace DQ_`var'=0 if DM_`var'
  5. }
(7 missing values generated)
(7 real changes made)
(7 missing values generated)
(7 real changes made)
(3 missing values generated)
(3 real changes made)
(3 missing values generated)
(3 real changes made)
(17 missing values generated)
(17 real changes made)
(17 missing values generated)
(17 real changes made)
(17 missing values generated)
(17 real changes made)
(2 missing values generated)
(2 real changes made)
(2 missing values generated)
(2 real changes made)
(2 missing values generated)
(2 real changes made)
(2 missing values generated)
(2 real changes made)
(2 missing values generated)
(2 real changes made)
(2 missing values generated)
(2 real changes made)
(2 missing values generated)
(2 real changes made)
(2 missing values generated)
(2 real changes made)
(0 real changes made)
(0 real changes made)
(0 real changes made)
(0 real changes made)

. 
. ** Dummy out cognition index
. gen CQ_cognix_pca = cognix_pca
(490 missing values generated)

. gen byte CM_cognix_pca =missing(cognix_pca)

. replace CQ_cognix_pca=0 if CM_cognix_pca
(490 real changes made)

. 
. ** Row 8: "Include observations with missing demographics (dummied out)"
. reg logspread    any_complexity consequence cognix_pca $exp_controls DQ_* DM_* if ~miss_cognix_pca, robust
note: DM_agesq omitted because of collinearity
note: DM_nhother omitted because of collinearity
note: DM_hispanic omitted because of collinearity
note: DM_ed_dropout omitted because of collinearity
note: DM_ed_hschool omitted because of collinearity
note: DM_ed_college omitted because of collinearity
note: DM_ed_graduat omitted because of collinearity
note: DM_hinc_25_50 omitted because of collinearity
note: DM_hinc_50_75 omitted because of collinearity
note: DM_hinc_75_100 omitted because of collinearity
note: DM_hinc_ge100 omitted because of collinearity
note: DM_hhsiz_2 omitted because of collinearity
note: DM_hhsiz_3 omitted because of collinearity
note: DM_hhsiz_4p omitted because of collinearity
note: DM_anykids omitted because of collinearity

Linear regression                               Number of obs     =      4,081
                                                F(34, 4044)       =          .
                                                Prob > F          =          .
                                                R-squared         =     0.1585
                                                Root MSE          =     1.9605

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1302384   .0647095     2.01   0.044     .0033722    .2571046
    consequence |  -.1387305   .0615175    -2.26   0.024    -.2593386   -.0181224
     cognix_pca |  -.7908737   .0427247   -18.51   0.000    -.8746376   -.7071099
     sell_first |   .1602395   .0615459     2.60   0.009     .0395756    .2809034
ls_startvalue_2 |   .0659818   .0759533     0.87   0.385    -.0829285     .214892
ls_startvalue_3 |  -.0025307   .0747775    -0.03   0.973    -.1491358    .1440745
       ls_first |   .0283755   .0615191     0.46   0.645    -.0922358    .1489867
   ss_benefit_2 |   .1061538   .0869462     1.22   0.222    -.0643087    .2766162
   ss_benefit_3 |   .0563081   .0841116     0.67   0.503    -.1085971    .2212133
   ss_benefit_4 |   .1515811   .0866396     1.75   0.080    -.0182801    .3214424
vignette_name_2 |   .1049809   .0854388     1.23   0.219    -.0625262    .2724881
vignette_name_3 |   .0779424   .0876003     0.89   0.374    -.0938024    .2496872
vignette_name_4 |  -.0202249     .08491    -0.24   0.812    -.1866952    .1462453
         DQ_age |   .0254947   .0129604     1.97   0.049     .0000852    .0509042
       DQ_agesq |  -.0153094   .0128873    -1.19   0.235    -.0405756    .0099569
      DQ_female |   .0866375   .0662091     1.31   0.191    -.0431689    .2164439
     DQ_married |   .0963439   .0757629     1.27   0.204    -.0521932    .2448809
     DQ_nhblack |   .0247119   .1419679     0.17   0.862    -.2536234    .3030472
     DQ_nhother |   .0480272   .1215521     0.40   0.693    -.1902819    .2863364
    DQ_hispanic |   .0891666   .1250026     0.71   0.476    -.1559073    .3342405
  DQ_ed_dropout |   -.041699   .1782222    -0.23   0.815    -.3911127    .3077146
  DQ_ed_hschool |   .0302406    .092918     0.33   0.745    -.1519298    .2124111
  DQ_ed_college |   .0105063   .0836515     0.13   0.900    -.1534967    .1745094
  DQ_ed_graduat |   .0781685   .0953786     0.82   0.413    -.1088261    .2651631
  DQ_hinc_25_50 |   .0938677   .1164064     0.81   0.420     -.134353    .3220884
  DQ_hinc_50_75 |  -.1685876      .1158    -1.46   0.146    -.3956194    .0584442
 DQ_hinc_75_100 |  -.0595895   .1298504    -0.46   0.646    -.3141678    .1949889
  DQ_hinc_ge100 |  -.2644079   .1098037    -2.41   0.016    -.4796837   -.0491321
     DQ_hhsiz_2 |  -.0223036   .0952271    -0.23   0.815    -.2090012    .1643941
     DQ_hhsiz_3 |   .1565557    .130157     1.20   0.229    -.0986238    .4117352
    DQ_hhsiz_4p |   .1857478   .1444027     1.29   0.198     -.097361    .4688566
     DQ_anykids |  -.1818119   .1049172    -1.73   0.083    -.3875075    .0238836
         DM_age |   2.016839   .9534873     2.12   0.034     .1474792      3.8862
       DM_agesq |          0  (omitted)
      DM_female |   -1.10088   1.056059    -1.04   0.297    -3.171337    .9695777
     DM_married |  -.7373706   .1546681    -4.77   0.000    -1.040605    -.434136
     DM_nhblack |   .5260863   .5002711     1.05   0.293    -.4547207    1.506893
     DM_nhother |          0  (omitted)
    DM_hispanic |          0  (omitted)
  DM_ed_dropout |          0  (omitted)
  DM_ed_hschool |          0  (omitted)
  DM_ed_college |          0  (omitted)
  DM_ed_graduat |          0  (omitted)
  DM_hinc_25_50 |          0  (omitted)
  DM_hinc_50_75 |          0  (omitted)
 DM_hinc_75_100 |          0  (omitted)
  DM_hinc_ge100 |          0  (omitted)
     DM_hhsiz_2 |          0  (omitted)
     DM_hhsiz_3 |          0  (omitted)
    DM_hhsiz_4p |          0  (omitted)
     DM_anykids |          0  (omitted)
          _cons |   1.103315   .3406912     3.24   0.001     .4353727    1.771257
---------------------------------------------------------------------------------

. 
. ** Row 9: "Include observations with missing cognition index (dummied out)"
. reg logspread    any_complexity consequence CQ_* CM_* $exp_controls $demographics if ~miss_anydemographic, robust

Linear regression                               Number of obs     =      4,528
                                                F(33, 4494)       =      24.10
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1422
                                                Root MSE          =     1.9876

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1179408     .06237     1.89   0.059     -.004335    .2402167
    consequence |   -.120164   .0591626    -2.03   0.042    -.2361519   -.0041761
  CQ_cognix_pca |  -.7411884   .0412921   -17.95   0.000    -.8221412   -.6602355
  CM_cognix_pca |   .2223973   .1066885     2.08   0.037     .0132353    .4315593
     sell_first |   .1903368   .0591181     3.22   0.001     .0744363    .3062373
ls_startvalue_2 |   .0847349    .073163     1.16   0.247    -.0587006    .2281704
ls_startvalue_3 |   .0495806   .0717749     0.69   0.490    -.0911335    .1902948
       ls_first |   .0347724    .059135     0.59   0.557    -.0811612    .1507061
   ss_benefit_2 |   .1261517   .0829185     1.52   0.128    -.0364095    .2887128
   ss_benefit_3 |    .107985     .08118     1.33   0.184    -.0511677    .2671377
   ss_benefit_4 |   .1754189   .0831637     2.11   0.035     .0123771    .3384607
vignette_name_2 |   .0566253   .0817799     0.69   0.489    -.1037035     .216954
vignette_name_3 |   .0961463   .0845463     1.14   0.256    -.0696061    .2618987
vignette_name_4 |  -.0116399   .0814375    -0.14   0.886    -.1712974    .1480176
            age |   .0227982   .0123847     1.84   0.066    -.0014819    .0470783
          agesq |  -.0133918    .012366    -1.08   0.279    -.0376353    .0108517
         female |   .1122689   .0633421     1.77   0.076    -.0119128    .2364507
        married |   .0685194   .0727029     0.94   0.346     -.074014    .2110528
        nhblack |   .1183743   .1318531     0.90   0.369    -.1401227    .3768712
        nhother |   .0841013   .1142579     0.74   0.462    -.1399004    .3081031
       hispanic |   .1479293   .1171962     1.26   0.207    -.0818329    .3776915
     ed_dropout |   .0577707   .1681141     0.34   0.731    -.2718157    .3873571
     ed_hschool |   .0782985    .088092     0.89   0.374    -.0944053    .2510022
     ed_college |  -.0182306   .0806637    -0.23   0.821    -.1763712      .13991
     ed_graduat |  -.0396478    .091522    -0.43   0.665     -.219076    .1397804
     hinc_25_50 |   .1175571    .110691     1.06   0.288    -.0994517    .3345659
     hinc_50_75 |  -.1352518   .1096964    -1.23   0.218    -.3503108    .0798072
    hinc_75_100 |  -.0887706    .123571    -0.72   0.473    -.3310306    .1534894
     hinc_ge100 |  -.2511707   .1046873    -2.40   0.016    -.4564093   -.0459321
        hhsiz_2 |  -.0332827   .0911472    -0.37   0.715    -.2119759    .1454106
        hhsiz_3 |   .1521884   .1241264     1.23   0.220    -.0911604    .3955372
       hhsiz_4p |   .1673883   .1374476     1.22   0.223    -.1020767    .4368533
        anykids |  -.1894469   .1002258    -1.89   0.059    -.3859387    .0070449
          _cons |    1.11976   .3248115     3.45   0.001     .4829701    1.756551
---------------------------------------------------------------------------------

. 
. ** Row 10: "Include observations with any missing values (dummied out)"
. reg logspread    any_complexity consequence CQ_* CM_* $exp_controls DQ_* DM_*, robust
note: DM_agesq omitted because of collinearity
note: DM_nhother omitted because of collinearity
note: DM_hispanic omitted because of collinearity
note: DM_ed_hschool omitted because of collinearity
note: DM_ed_college omitted because of collinearity
note: DM_ed_graduat omitted because of collinearity
note: DM_hinc_50_75 omitted because of collinearity
note: DM_hinc_75_100 omitted because of collinearity
note: DM_hinc_ge100 omitted because of collinearity
note: DM_hhsiz_2 omitted because of collinearity
note: DM_hhsiz_3 omitted because of collinearity
note: DM_hhsiz_4p omitted because of collinearity
note: DM_anykids omitted because of collinearity

Linear regression                               Number of obs     =      4,552
                                                F(37, 4512)       =          .
                                                Prob > F          =          .
                                                R-squared         =     0.1441
                                                Root MSE          =     1.9874

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1165629   .0622709     1.87   0.061    -.0055185    .2386442
    consequence |  -.1184843   .0590691    -2.01   0.045    -.2342886   -.0026799
  CQ_cognix_pca |  -.7444665   .0412673   -18.04   0.000    -.8253705   -.6635624
  CM_cognix_pca |   .2218557   .1067431     2.08   0.038     .0125869    .4311245
     sell_first |   .1857037   .0590248     3.15   0.002     .0699861    .3014212
ls_startvalue_2 |    .087043   .0730399     1.19   0.233    -.0561509     .230237
ls_startvalue_3 |    .048779   .0716304     0.68   0.496    -.0916517    .1892097
       ls_first |   .0335001   .0590614     0.57   0.571    -.0822891    .1492894
   ss_benefit_2 |   .1202187   .0828556     1.45   0.147    -.0422189    .2826563
   ss_benefit_3 |   .1070694   .0810896     1.32   0.187    -.0519059    .2660447
   ss_benefit_4 |   .1610642   .0830116     1.94   0.052    -.0016793    .3238077
vignette_name_2 |   .0488662   .0816518     0.60   0.550    -.1112113    .2089437
vignette_name_3 |    .087676   .0843927     1.04   0.299    -.0777751    .2531271
vignette_name_4 |  -.0186692    .081242    -0.23   0.818    -.1779433    .1406049
         DQ_age |   .0234391   .0123683     1.90   0.058    -.0008087     .047687
       DQ_agesq |  -.0139345   .0123527    -1.13   0.259    -.0381518    .0102828
      DQ_female |   .1133244   .0632505     1.79   0.073    -.0106775    .2373262
     DQ_married |   .0683783   .0725228     0.94   0.346    -.0738019    .2105585
     DQ_nhblack |   .1158347   .1319744     0.88   0.380    -.1428997    .3745692
     DQ_nhother |   .0841071   .1143074     0.74   0.462    -.1399913    .3082056
    DQ_hispanic |   .1548194    .117049     1.32   0.186    -.0746539    .3842927
  DQ_ed_dropout |   .0717562   .1684508     0.43   0.670    -.2584898    .4020023
  DQ_ed_hschool |   .0756392   .0880296     0.86   0.390    -.0969421    .2482204
  DQ_ed_college |  -.0158197    .080504    -0.20   0.844    -.1736468    .1420075
  DQ_ed_graduat |  -.0387269   .0914283    -0.42   0.672    -.2179711    .1405173
  DQ_hinc_25_50 |   .1115077   .1104031     1.01   0.313    -.1049364    .3279518
  DQ_hinc_50_75 |  -.1367765   .1095735    -1.25   0.212    -.3515943    .0780413
 DQ_hinc_75_100 |  -.0919231   .1234246    -0.74   0.456    -.3338958    .1500495
  DQ_hinc_ge100 |  -.2568345   .1045477    -2.46   0.014    -.4617991   -.0518698
     DQ_hhsiz_2 |  -.0309591   .0910182    -0.34   0.734    -.2093993    .1474811
     DQ_hhsiz_3 |   .1599394   .1237917     1.29   0.196    -.0827531    .4026318
    DQ_hhsiz_4p |    .171872   .1367958     1.26   0.209    -.0963148    .4400588
     DQ_anykids |  -.1944923   .0996432    -1.95   0.051    -.3898419    .0008572
         DM_age |   1.910078    .949463     2.01   0.044     .0486654    3.771491
       DM_agesq |          0  (omitted)
      DM_female |  -1.635746   .1688415    -9.69   0.000    -1.966758   -1.304734
     DM_married |  -.8569176   .1486934    -5.76   0.000     -1.14843   -.5654057
     DM_nhblack |   .5347077   .4969448     1.08   0.282    -.4395475    1.508963
     DM_nhother |          0  (omitted)
    DM_hispanic |          0  (omitted)
  DM_ed_dropout |   1.217124   1.153643     1.06   0.291    -1.044581     3.47883
  DM_ed_hschool |          0  (omitted)
  DM_ed_college |          0  (omitted)
  DM_ed_graduat |          0  (omitted)
  DM_hinc_25_50 |  -2.051233   .5025197    -4.08   0.000    -3.036417   -1.066048
  DM_hinc_50_75 |          0  (omitted)
 DM_hinc_75_100 |          0  (omitted)
  DM_hinc_ge100 |          0  (omitted)
     DM_hhsiz_2 |          0  (omitted)
     DM_hhsiz_3 |          0  (omitted)
    DM_hhsiz_4p |          0  (omitted)
     DM_anykids |          0  (omitted)
          _cons |   1.116295   .3237687     3.45   0.001     .4815501    1.751041
---------------------------------------------------------------------------------

. 
. ** double check that N matches the count below
. count if ~miss_spread
  4,552

. assert e(sample)==~miss_spread 

. 
. ** Row 11: "Exclude Native American and LA County oversamples"
. reg logspread    any_complexity consequence cognix_pca $exp_controls $demographics if basesample & nationalsample, robust

Linear regression                               Number of obs     =      3,704
                                                F(32, 3671)       =      23.33
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1632
                                                Root MSE          =     1.9498

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1127289   .0679952     1.66   0.097    -.0205832     .246041
    consequence |  -.1686902   .0641453    -2.63   0.009    -.2944541   -.0429263
     cognix_pca |  -.8081782   .0443241   -18.23   0.000    -.8950806   -.7212759
     sell_first |    .164993   .0642526     2.57   0.010     .0390186    .2909674
ls_startvalue_2 |   .0543311   .0790581     0.69   0.492     -.100671    .2093332
ls_startvalue_3 |  -.0045397   .0781054    -0.06   0.954    -.1576739    .1485945
       ls_first |   .0295843   .0640933     0.46   0.644    -.0960778    .1552463
   ss_benefit_2 |   .1223145   .0914081     1.34   0.181    -.0569012    .3015301
   ss_benefit_3 |    .100089    .087653     1.14   0.254    -.0717643    .2719423
   ss_benefit_4 |   .1615541   .0903387     1.79   0.074     -.015565    .3386732
vignette_name_2 |   .1597016   .0888702     1.80   0.072    -.0145382    .3339414
vignette_name_3 |   .1379244   .0914974     1.51   0.132    -.0414664    .3173151
vignette_name_4 |   .0374168   .0882051     0.42   0.671    -.1355192    .2103527
            age |   .0218964   .0134542     1.63   0.104    -.0044821    .0482748
          agesq |  -.0119035   .0132947    -0.90   0.371    -.0379692    .0141621
         female |    .063277   .0685935     0.92   0.356    -.0712082    .1977621
        married |   .0877228   .0799934     1.10   0.273    -.0691131    .2445587
        nhblack |    .107336   .1491689     0.72   0.472    -.1851262    .3997981
        nhother |   .1093424   .1389173     0.79   0.431    -.1630204    .3817052
       hispanic |   .1933157   .1469263     1.32   0.188    -.0947496     .481381
     ed_dropout |  -.0738682   .1894172    -0.39   0.697    -.4452414    .2975051
     ed_hschool |   .0454642    .097132     0.47   0.640    -.1449738    .2359022
     ed_college |  -.0149876   .0872334    -0.17   0.864    -.1860184    .1560432
     ed_graduat |   .0723564   .0983227     0.74   0.462    -.1204161    .2651288
     hinc_25_50 |   .1758383   .1239862     1.42   0.156    -.0672504    .4189269
     hinc_50_75 |  -.1322479   .1219237    -1.08   0.278    -.3712928     .106797
    hinc_75_100 |  -.0125109   .1363258    -0.09   0.927    -.2797927    .2547709
     hinc_ge100 |  -.2088995   .1155805    -1.81   0.071    -.4355079    .0177089
        hhsiz_2 |  -.0173115   .0986971    -0.18   0.861     -.210818     .176195
        hhsiz_3 |   .1719352   .1355939     1.27   0.205    -.0939115     .437782
       hhsiz_4p |   .2072334   .1528403     1.36   0.175    -.0924268    .5068936
        anykids |   -.141747   .1107531    -1.28   0.201    -.3588908    .0753968
          _cons |   1.108164   .3572765     3.10   0.002     .4076843    1.808644
---------------------------------------------------------------------------------

. 
. 
. 
. 
. ** Panel C: Different controls
. ** ------------------------------------
. 
. ** Row 12: "No Cognition Controls"
. reg logspread    any_complexity consequence            $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(31, 4028)       =      12.21
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0825
                                                Root MSE          =     2.0446

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .0871538   .0675366     1.29   0.197    -.0452554    .2195629
    consequence |  -.1591144   .0642751    -2.48   0.013    -.2851292   -.0330996
     sell_first |    .129214   .0643472     2.01   0.045     .0030578    .2553701
ls_startvalue_2 |    .033813   .0791032     0.43   0.669     -.121273     .188899
ls_startvalue_3 |  -.0472465   .0783751    -0.60   0.547    -.2009051    .1064122
       ls_first |   .0588945   .0642435     0.92   0.359    -.0670582    .1848473
   ss_benefit_2 |   .1435612   .0904029     1.59   0.112    -.0336785    .3208008
   ss_benefit_3 |   .0716091   .0876972     0.82   0.414     -.100326    .2435441
   ss_benefit_4 |   .1673411   .0909569     1.84   0.066    -.0109848    .3456671
vignette_name_2 |    .101395   .0888027     1.14   0.254    -.0727073    .2754973
vignette_name_3 |   .0532422   .0911155     0.58   0.559    -.1253946    .2318791
vignette_name_4 |  -.0306794   .0895165    -0.34   0.732    -.2061813    .1448226
            age |   .0093695   .0136283     0.69   0.492    -.0173495    .0360886
          agesq |  -.0026825   .0136021    -0.20   0.844    -.0293501    .0239851
         female |   .4031203   .0670431     6.01   0.000     .2716787    .5345619
        married |   .0865948     .07895     1.10   0.273    -.0681909    .2413804
        nhblack |   .7557525   .1377846     5.49   0.000     .4856185    1.025887
        nhother |   .2778327   .1275963     2.18   0.030     .0276734    .5279921
       hispanic |   .5183105   .1279512     4.05   0.000     .2674554    .7691657
     ed_dropout |   .5027741   .1776915     2.83   0.005     .1544005    .8511477
     ed_hschool |   .3450643   .0941659     3.66   0.000     .1604471    .5296814
     ed_college |  -.3918924   .0843982    -4.64   0.000    -.5573596   -.2264252
     ed_graduat |  -.4495651   .0960364    -4.68   0.000    -.6378497   -.2612806
     hinc_25_50 |  -.0313384     .12121    -0.26   0.796     -.268977    .2063003
     hinc_50_75 |  -.3960074   .1186951    -3.34   0.001    -.6287155   -.1632993
    hinc_75_100 |  -.2848167   .1340745    -2.12   0.034    -.5476769   -.0219566
     hinc_ge100 |  -.5613872   .1129334    -4.97   0.000    -.7827992   -.3399752
        hhsiz_2 |  -.1317308   .0989785    -1.33   0.183    -.3257833    .0623218
        hhsiz_3 |    .028216   .1353443     0.21   0.835    -.2371337    .2935657
       hhsiz_4p |   .0252211   .1497468     0.17   0.866    -.2683654    .3188076
        anykids |  -.1445168   .1096722    -1.32   0.188    -.3595349    .0705013
          _cons |   1.687352   .3583787     4.71   0.000     .9847321    2.389973
---------------------------------------------------------------------------------

. 
. ** Row 13: "No Demographic Controls"
. reg logspread    any_complexity consequence cognix_pca $exp_controls               if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(13, 4046)       =      52.71
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1465
                                                Root MSE          =     1.9675

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1372385   .0645476     2.13   0.034     .0106896    .2637874
    consequence |  -.1401168   .0618885    -2.26   0.024    -.2614522   -.0187813
     cognix_pca |  -.8022804   .0316034   -25.39   0.000    -.8642404   -.7403204
     sell_first |   .1647436   .0619323     2.66   0.008     .0433222     .286165
ls_startvalue_2 |     .06113    .075976     0.80   0.421    -.0878249    .2100848
ls_startvalue_3 |   .0136391   .0751617     0.18   0.856    -.1337192    .1609975
       ls_first |   .0421793   .0618227     0.68   0.495    -.0790272    .1633859
   ss_benefit_2 |   .1069014   .0871503     1.23   0.220    -.0639612     .277764
   ss_benefit_3 |   .0688483   .0842175     0.82   0.414    -.0962643     .233961
   ss_benefit_4 |   .1731118   .0869422     1.99   0.047     .0026572    .3435663
vignette_name_2 |   .1132578   .0859084     1.32   0.187      -.05517    .2816857
vignette_name_3 |   .0827615   .0882784     0.94   0.349    -.0903127    .2558357
vignette_name_4 |  -.0169665   .0850449    -0.20   0.842    -.1837014    .1497684
          _cons |   1.938982   .1120224    17.31   0.000     1.719356    2.158607
---------------------------------------------------------------------------------

. 
. ** Row 14: "No Secondary Experimental Controls"
. reg logspread    any_complexity consequence cognix_pca               $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(22, 4037)       =      33.14
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1534
                                                Root MSE          =     1.9618

--------------------------------------------------------------------------------
               |               Robust
     logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
any_complexity |   .1246668   .0648144     1.92   0.054    -.0024051    .2517388
   consequence |  -.1382234   .0616243    -2.24   0.025     -.259041   -.0174058
    cognix_pca |  -.7839514      .0428   -18.32   0.000     -.867863   -.7000398
           age |   .0237142   .0129348     1.83   0.067    -.0016452    .0490735
         agesq |  -.0136497    .012847    -1.06   0.288    -.0388369    .0115375
        female |   .0827258   .0663496     1.25   0.213     -.047356    .2128075
       married |   .0958755   .0757187     1.27   0.206     -.052575     .244326
       nhblack |   .0324218   .1419981     0.23   0.819    -.2459728    .3108165
       nhother |   .0627667   .1211723     0.52   0.604     -.174798    .3003313
      hispanic |   .0836906   .1253243     0.67   0.504    -.1620143    .3293954
    ed_dropout |  -.0454376   .1779917    -0.26   0.799    -.3943996    .3035244
    ed_hschool |   .0336542   .0930051     0.36   0.717    -.1486872    .2159955
    ed_college |    .006067   .0836756     0.07   0.942    -.1579834    .1701174
    ed_graduat |   .0682161   .0952072     0.72   0.474    -.1184425    .2548746
    hinc_25_50 |   .1047924   .1169154     0.90   0.370    -.1244262     .334011
    hinc_50_75 |  -.1633842   .1163364    -1.40   0.160    -.3914677    .0646994
   hinc_75_100 |  -.0587691   .1301049    -0.45   0.652    -.3138464    .1963082
    hinc_ge100 |  -.2527527   .1099464    -2.30   0.022    -.4683084   -.0371969
       hhsiz_2 |    -.02685    .095202    -0.28   0.778    -.2134984    .1597984
       hhsiz_3 |    .152384   .1303205     1.17   0.242     -.103116    .4078841
      hhsiz_4p |   .1735735   .1448862     1.20   0.231    -.1104835    .4576304
       anykids |  -.1793623   .1056449    -1.70   0.090    -.3864845    .0277599
         _cons |   1.382383   .3270704     4.23   0.000     .7411451    2.023622
--------------------------------------------------------------------------------

. 
. 
. 
. 
. ** Panel D: Adjustments to the outcome variable
. ** ---------------------------------------------------
. 
. ** Row 15: "Buy and sell valuations topcoded at $100,000"
. reg logspread_100k any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      21.21
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1427
                                                Root MSE          =     1.6285

---------------------------------------------------------------------------------
                |               Robust
 logspread_100k |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1111977   .0538035     2.07   0.039      .005713    .2166824
    consequence |   -.108341   .0512423    -2.11   0.035    -.2088043   -.0078778
     cognix_pca |  -.6104148   .0352084   -17.34   0.000    -.6794427   -.5413868
     sell_first |   .0384236   .0513519     0.75   0.454    -.0622546    .1391018
ls_startvalue_2 |  -.0087262   .0637311    -0.14   0.891    -.1336743    .1162219
ls_startvalue_3 |  -.1061172    .061978    -1.71   0.087    -.2276283    .0153939
       ls_first |   .0457435   .0512321     0.89   0.372    -.0546998    .1461868
   ss_benefit_2 |    .160804   .0731815     2.20   0.028     .0173277    .3042802
   ss_benefit_3 |   .0878748   .0700427     1.25   0.210    -.0494477    .2251973
   ss_benefit_4 |   .1559158   .0715561     2.18   0.029     .0156262    .2962054
vignette_name_2 |    .098694   .0720407     1.37   0.171    -.0425456    .2399335
vignette_name_3 |    .005407   .0727651     0.07   0.941    -.1372529    .1480669
vignette_name_4 |   -.075758   .0706353    -1.07   0.284    -.2142423    .0627264
            age |   .0236274   .0105703     2.24   0.025     .0029037    .0443512
          agesq |  -.0142618    .010552    -1.35   0.177    -.0349496    .0064259
         female |   .0647425   .0550406     1.18   0.240    -.0431675    .1726525
        married |   .0755239   .0626126     1.21   0.228    -.0472315    .1982793
        nhblack |   .0286143    .117426     0.24   0.807    -.2016056    .2588342
        nhother |    .038234   .0994358     0.38   0.701    -.1567152    .2331832
       hispanic |    .124244   .1019308     1.22   0.223    -.0755967    .3240848
     ed_dropout |  -.1029857   .1435157    -0.72   0.473    -.3843559    .1783845
     ed_hschool |   .0006152   .0776109     0.01   0.994     -.151545    .1527754
     ed_college |  -.0052542   .0690873    -0.08   0.939    -.1407034    .1301951
     ed_graduat |   .0400269   .0795779     0.50   0.615    -.1159899    .1960436
     hinc_25_50 |   .0581205   .0958238     0.61   0.544    -.1297472    .2459882
     hinc_50_75 |  -.1639203   .0953437    -1.72   0.086    -.3508466     .023006
    hinc_75_100 |  -.1092864   .1064138    -1.03   0.304    -.3179163    .0993435
     hinc_ge100 |  -.2110536   .0904628    -2.33   0.020    -.3884107   -.0336965
        hhsiz_2 |  -.0310982   .0789741    -0.39   0.694    -.1859312    .1237348
        hhsiz_3 |   .1267991     .10847     1.17   0.242    -.0858622    .3394603
       hhsiz_4p |   .1577716   .1207596     1.31   0.191     -.078984    .3945271
        anykids |  -.1714536   .0883181    -1.94   0.052    -.3446058    .0016987
          _cons |   1.005545   .2785495     3.61   0.000     .4594337    1.551656
---------------------------------------------------------------------------------

. 
. 
. ** Row 16: "Topcoding spread at the 90th percentile"
. sum logspread if basesample, d

                         Log Spread
-------------------------------------------------------------
      Percentiles      Smallest
 1%            0              0
 5%            0              0
10%            0              0       Obs               4,060
25%     .4353181              0       Sum of Wgt.       4,060

50%     1.547563                      Mean           2.214562
                        Largest       Std. Dev.      2.126367
75%     3.563716       8.006368
90%     5.393628       8.006368       Variance       4.521437
95%      6.55108       8.006368       Skewness       1.006302
99%     8.006368       8.006368       Kurtosis       3.133193

. gen logspread_top90 = logspread
(44 missing values generated)

. replace logspread_top90 =  r(p90) if logspread>r(p90) & logspread<.
(465 real changes made)

. reg logspread_top90 any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      26.63
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1641
                                                Root MSE          =     1.6766

---------------------------------------------------------------------------------
                |               Robust
logspread_top90 |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1186714   .0553798     2.14   0.032     .0100963    .2272465
    consequence |  -.1061761   .0526871    -2.02   0.044    -.2094719   -.0028803
     cognix_pca |   -.689534   .0357264   -19.30   0.000    -.7595775   -.6194904
     sell_first |   .2128473   .0527432     4.04   0.000     .1094413    .3162532
ls_startvalue_2 |   .0412873   .0646634     0.64   0.523    -.0854888    .1680634
ls_startvalue_3 |   .0091298   .0645737     0.14   0.888    -.1174703    .1357299
       ls_first |   .0378255   .0527247     0.72   0.473    -.0655442    .1411951
   ss_benefit_2 |   .0885066   .0751243     1.18   0.239    -.0587785    .2357918
   ss_benefit_3 |   .0455847   .0725638     0.63   0.530    -.0966805    .1878498
   ss_benefit_4 |   .1206961    .073892     1.63   0.102    -.0241732    .2655653
vignette_name_2 |    .107964   .0736343     1.47   0.143    -.0363999     .252328
vignette_name_3 |    .063978   .0751818     0.85   0.395    -.0834199    .2113759
vignette_name_4 |   -.030026   .0725799    -0.41   0.679    -.1723228    .1122708
            age |   .0237676   .0105665     2.25   0.025     .0030514    .0444837
          agesq |  -.0149166   .0104601    -1.43   0.154    -.0354241    .0055909
         female |   .0748914   .0565041     1.33   0.185    -.0358879    .1856706
        married |   .1013436   .0651392     1.56   0.120    -.0263652    .2290524
        nhblack |   .0189078   .1163985     0.16   0.871    -.2092977    .2471133
        nhother |      .0351   .1035471     0.34   0.735    -.1679096    .2381095
       hispanic |   .0993831    .105532     0.94   0.346    -.1075179    .3062842
     ed_dropout |  -.1271242   .1432158    -0.89   0.375    -.4079064    .1536581
     ed_hschool |   .0229789   .0790821     0.29   0.771    -.1320658    .1780237
     ed_college |  -.0235458   .0722404    -0.33   0.744    -.1651771    .1180854
     ed_graduat |    .043198   .0823566     0.52   0.600    -.1182664    .2046625
     hinc_25_50 |   .0826107   .0966303     0.85   0.393    -.1068381    .2720595
     hinc_50_75 |  -.1156875   .0987965    -1.17   0.242    -.3093833    .0780084
    hinc_75_100 |  -.0524823   .1096323    -0.48   0.632    -.2674223    .1624576
     hinc_ge100 |  -.2291704   .0912982    -2.51   0.012    -.4081654   -.0501754
        hhsiz_2 |  -.0242488   .0811517    -0.30   0.765     -.183351    .1348534
        hhsiz_3 |   .1016665   .1101812     0.92   0.356    -.1143495    .3176826
       hhsiz_4p |   .0919969   .1225516     0.75   0.453    -.1482721     .332266
        anykids |  -.0937975    .090065    -1.04   0.298    -.2703746    .0827797
          _cons |    1.02724   .2836688     3.62   0.000     .4710918    1.583387
---------------------------------------------------------------------------------

. 
. ** Row 17: "Bottomcoding buy and sell valuations at $1000"
. gen logspread_1k = abs(log((max(1000,sellpricesup)+max(1000,sellpriceinf))/2)   ///
>                      - log((max(1000, buypricesup)+max(1000, buypriceinf))/2))  if logspread<.
(44 missing values generated)

. reg logspread_1k any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      22.85
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1501
                                                Root MSE          =     1.7306

---------------------------------------------------------------------------------
                |               Robust
   logspread_1k |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |    .103546   .0572263     1.81   0.070    -.0086491    .2157412
    consequence |  -.0975581   .0543796    -1.79   0.073    -.2041722     .009056
     cognix_pca |  -.6820514   .0375768   -18.15   0.000    -.7557227   -.6083801
     sell_first |   .1853741   .0543938     3.41   0.001     .0787322     .292016
ls_startvalue_2 |    .089062   .0658349     1.35   0.176    -.0400108    .2181348
ls_startvalue_3 |   .1266504   .0669251     1.89   0.059    -.0045598    .2578606
       ls_first |   .0086205   .0543729     0.16   0.874    -.0979805    .1152216
   ss_benefit_2 |   .0486165   .0765927     0.63   0.526    -.1015476    .1987805
   ss_benefit_3 |   .0183202   .0746867     0.25   0.806    -.1281071    .1647474
   ss_benefit_4 |   .1000491   .0771235     1.30   0.195    -.0511556    .2512537
vignette_name_2 |   .0801918   .0752862     1.07   0.287    -.0674109    .2277944
vignette_name_3 |   .0728221   .0775811     0.94   0.348    -.0792797    .2249239
vignette_name_4 |  -.0130632   .0752101    -0.17   0.862    -.1605167    .1343903
            age |   .0199594   .0112725     1.77   0.077     -.002141    .0420599
          agesq |  -.0120149   .0111456    -1.08   0.281    -.0338665    .0098368
         female |   .0664951   .0584887     1.14   0.256    -.0481752    .1811654
        married |   .0792605   .0678873     1.17   0.243    -.0538362    .2123572
        nhblack |   .0041518   .1249602     0.03   0.973    -.2408393     .249143
        nhother |   .0437472   .1088653     0.40   0.688     -.169689    .2571834
       hispanic |   .0762921   .1109175     0.69   0.492    -.1411676    .2937519
     ed_dropout |  -.0824472   .1556878    -0.53   0.596    -.3876815    .2227871
     ed_hschool |   .0317726    .081831     0.39   0.698    -.1286614    .1922065
     ed_college |   .0202123   .0741736     0.27   0.785     -.125209    .1656336
     ed_graduat |   .0917763   .0838695     1.09   0.274    -.0726543    .2562068
     hinc_25_50 |    .052737   .1025475     0.51   0.607    -.1483128    .2537867
     hinc_50_75 |   -.160795   .1029708    -1.56   0.118    -.3626747    .0410847
    hinc_75_100 |  -.0771006   .1150566    -0.67   0.503    -.3026751    .1484739
     hinc_ge100 |  -.2492763   .0963185    -2.59   0.010    -.4381139   -.0604387
        hhsiz_2 |  -.0203591   .0841748    -0.24   0.809    -.1853883    .1446702
        hhsiz_3 |   .1373856   .1144824     1.20   0.230    -.0870632    .3618345
       hhsiz_4p |    .160544    .128145     1.25   0.210    -.0906911     .411779
        anykids |  -.1757204   .0934194    -1.88   0.060    -.3588742    .0074334
          _cons |    1.05935   .2999966     3.53   0.000     .4711902    1.647509
---------------------------------------------------------------------------------

. 
. ** Row 18: "Spread set to zero if spread ≤ 0.50"
. gen logspread_tol50 = logspread
(44 missing values generated)

. replace logspread_tol50 = 0 if logspread<=0.50
(716 real changes made)

. reg logspread_tol50 any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      24.60
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1592
                                                Root MSE          =     1.9944

---------------------------------------------------------------------------------
                |               Robust
logspread_tol50 |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1306686   .0659379     1.98   0.048     .0013939    .2599433
    consequence |  -.1418359   .0627063    -2.26   0.024     -.264775   -.0188969
     cognix_pca |  -.8087584   .0433934   -18.64   0.000    -.8938334   -.7236834
     sell_first |   .1777967   .0627192     2.83   0.005     .0548325     .300761
ls_startvalue_2 |   .0573355   .0774781     0.74   0.459    -.0945644    .2092355
ls_startvalue_3 |   .0014585   .0762093     0.02   0.985     -.147954     .150871
       ls_first |   .0301161   .0626857     0.48   0.631    -.0927824    .1530147
   ss_benefit_2 |   .1151236   .0885429     1.30   0.194    -.0584694    .2887167
   ss_benefit_3 |    .052606   .0858384     0.61   0.540    -.1156847    .2208967
   ss_benefit_4 |   .1719047    .088203     1.95   0.051    -.0010221    .3448315
vignette_name_2 |   .1177591   .0870497     1.35   0.176    -.0529064    .2884246
vignette_name_3 |   .0875984   .0893482     0.98   0.327    -.0875735    .2627703
vignette_name_4 |   -.012274   .0866019    -0.14   0.887    -.1820617    .1575136
            age |    .026162   .0131895     1.98   0.047     .0003032    .0520208
          agesq |  -.0158846   .0131144    -1.21   0.226    -.0415961    .0098268
         female |   .0885107   .0675499     1.31   0.190    -.0439245    .2209459
        married |   .0981161   .0772431     1.27   0.204    -.0533231    .2495552
        nhblack |   .0296578    .143477     0.21   0.836    -.2516365     .310952
        nhother |   .0455695   .1238006     0.37   0.713    -.1971482    .2882872
       hispanic |    .085641   .1269388     0.67   0.500    -.1632293    .3345114
     ed_dropout |  -.0667622    .180236    -0.37   0.711    -.4201244       .2866
     ed_hschool |   .0329623   .0944413     0.35   0.727    -.1521949    .2181195
     ed_college |  -.0026243   .0855663    -0.03   0.976    -.1703815    .1651329
     ed_graduat |    .082956   .0971144     0.85   0.393     -.107442     .273354
     hinc_25_50 |   .0987612   .1183268     0.83   0.404    -.1332247    .3307472
     hinc_50_75 |  -.1695261   .1177046    -1.44   0.150    -.4002922      .06124
    hinc_75_100 |   -.062366   .1320833    -0.47   0.637    -.3213223    .1965903
     hinc_ge100 |  -.2642164   .1115627    -2.37   0.018     -.482941   -.0454918
        hhsiz_2 |  -.0169423   .0970325    -0.17   0.861    -.2071797    .1732951
        hhsiz_3 |   .1530907   .1326347     1.15   0.248    -.1069466     .413128
       hhsiz_4p |   .1939337   .1474333     1.32   0.188     -.095117    .4829845
        anykids |  -.1844826   .1072606    -1.72   0.086    -.3947726    .0258075
          _cons |   1.021442   .3473454     2.94   0.003     .3404524    1.702431
---------------------------------------------------------------------------------

. di "A 0.50 log difference translates to a factor of: " exp(0.50)  
A 0.50 log difference translates to a factor of: 1.6487213

. 
. ** Row 19: "Spread set to zero if buy valuation > sell valuation"
. gen logspread_asym = logspread
(44 missing values generated)

. replace logspread_asym = 0 if logbuyprice>logsellprice & logbuyprice<.
(1,261 real changes made)

. reg logspread_asym any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      12.07
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0930
                                                Root MSE          =     2.0353

---------------------------------------------------------------------------------
                |               Robust
 logspread_asym |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1589171   .0672521     2.36   0.018     .0270658    .2907684
    consequence |   -.131678   .0640046    -2.06   0.040    -.2571623   -.0061936
     cognix_pca |  -.5370814   .0451712   -11.89   0.000     -.625642   -.4485208
     sell_first |   -.326859    .064027    -5.11   0.000    -.4523873   -.2013307
ls_startvalue_2 |   .0335027   .0782865     0.43   0.669    -.1199821    .1869876
ls_startvalue_3 |   .0028698   .0785536     0.04   0.971    -.1511387    .1568783
       ls_first |   .0253826    .064037     0.40   0.692    -.1001654    .1509306
   ss_benefit_2 |   .2906783   .0905744     3.21   0.001     .1131023    .4682542
   ss_benefit_3 |   .2223539   .0866147     2.57   0.010     .0525412    .3921666
   ss_benefit_4 |   .2009472   .0895124     2.24   0.025     .0254534    .3764411
vignette_name_2 |   .0919397   .0897417     1.02   0.306    -.0840036    .2678831
vignette_name_3 |  -.0614745   .0903596    -0.68   0.496    -.2386293    .1156802
vignette_name_4 |  -.1191726   .0884269    -1.35   0.178    -.2925382     .054193
            age |    .030387   .0139286     2.18   0.029     .0030793    .0576948
          agesq |  -.0162109   .0139612    -1.16   0.246    -.0435827    .0111608
         female |   .0847696   .0682138     1.24   0.214    -.0489673    .2185064
        married |    .096701   .0782259     1.24   0.216     -.056665     .250067
        nhblack |   .0283824   .1510779     0.19   0.851     -.267814    .3245787
        nhother |   .0398044   .1242189     0.32   0.749    -.2037334    .2833423
       hispanic |   .0418221   .1341339     0.31   0.755    -.2211545    .3047988
     ed_dropout |  -.0274941   .1869394    -0.15   0.883    -.3939987    .3390105
     ed_hschool |   .0447188   .0984133     0.45   0.650    -.1482258    .2376634
     ed_college |  -.0364942   .0851334    -0.43   0.668    -.2034027    .1304143
     ed_graduat |   .0335726   .0982436     0.34   0.733    -.1590392    .2261843
     hinc_25_50 |   .1420696   .1255965     1.13   0.258     -.104169    .3883081
     hinc_50_75 |  -.0994839    .120517    -0.83   0.409    -.3357638     .136796
    hinc_75_100 |  -.0744064   .1339823    -0.56   0.579     -.337086    .1882731
     hinc_ge100 |  -.1571979   .1144661    -1.37   0.170    -.3816147     .067219
        hhsiz_2 |   .0119997   .0984088     0.12   0.903    -.1809359    .2049353
        hhsiz_3 |   .2121665    .137079     1.55   0.122    -.0565841    .4809172
       hhsiz_4p |   .2833254   .1515852     1.87   0.062    -.0138654    .5805162
        anykids |  -.2662991   .1103224    -2.41   0.016     -.482592   -.0500062
          _cons |   .4219021   .3561278     1.18   0.236    -.2763054     1.12011
---------------------------------------------------------------------------------

. 
. 
. 
. 
. 
. 
. 
. // ***************************************************************************************
. //                      APX TABLE A09: Other Predictors of the Sell-Buy Spread
. // ***************************************************************************************
. 
. ** Row 1: "Cognition Index (Standardized)"
. reg logspread cognix_pca                                 , robust

Linear regression                               Number of obs     =      4,081
                                                F(1, 4079)        =     643.85
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1428
                                                Root MSE          =     1.9702

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  cognix_pca |  -.8038501   .0316797   -25.37   0.000    -.8659597   -.7417405
       _cons |   2.216756   .0308243    71.92   0.000     2.156323    2.277188
------------------------------------------------------------------------------

. 
. ** Row 2: "Decision-Making Competence, Framing Consistency (Standardized)"
. reg logspread admc_Sframe*                               , robust

Linear regression                               Number of obs     =      4,540
                                                F(1, 4538)        =     102.37
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0238
                                                Root MSE          =     2.1126

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
 admc_Sframe |  -.3302502   .0326397   -10.12   0.000    -.3942398   -.2662605
       _cons |   2.236579   .0313544    71.33   0.000     2.175109    2.298049
------------------------------------------------------------------------------

. 
. ** Row 3: "Decision-Making Competence, Time Conjunction (Standardized)"
. reg logspread admc_Stime*                                , robust

Linear regression                               Number of obs     =      4,540
                                                F(1, 4538)        =       5.14
                                                Prob > F          =     0.0235
                                                R-squared         =     0.0012
                                                Root MSE          =      2.137

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  admc_Stime |  -.0744623   .0328572    -2.27   0.023    -.1388784   -.0100462
       _cons |   2.236542   .0317151    70.52   0.000     2.174365    2.298719
------------------------------------------------------------------------------

. 
. ** Row 4: "Decision-Making Competence, Subset Consistency (Standardized)"
. reg logspread admc_Ssubset*                              , robust

Linear regression                               Number of obs     =      4,540
                                                F(1, 4538)        =      31.24
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0077
                                                Root MSE          =       2.13

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
admc_Ssubset |  -.1889582   .0338062    -5.59   0.000    -.2552348   -.1226815
       _cons |   2.237127   .0316168    70.76   0.000     2.175143    2.299112
------------------------------------------------------------------------------

. 
. ** Row 5: "Self-Assessed Knowledge about Social Security (Standardized)"
. reg logspread ssa_Sknowledge*                            , robust

Linear regression                               Number of obs     =      4,248
                                                F(1, 4246)        =      13.05
                                                Prob > F          =     0.0003
                                                R-squared         =     0.0032
                                                Root MSE          =     2.1349

--------------------------------------------------------------------------------
               |               Robust
     logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
ssa_Sknowledge |  -.1205149   .0333562    -3.61   0.000    -.1859106   -.0551193
         _cons |   2.216848   .0327552    67.68   0.000     2.152631    2.281066
--------------------------------------------------------------------------------

. 
. ** Row 6: "Social Security Literacy Score (Standardized)"
. reg logspread ssa_Sliteracy*                             , robust

Linear regression                               Number of obs     =      4,257
                                                F(1, 4255)        =      80.33
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0189
                                                Root MSE          =     2.1135

-------------------------------------------------------------------------------
              |               Robust
    logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
ssa_Sliteracy |  -.2937506   .0327746    -8.96   0.000    -.3580059   -.2294953
        _cons |    2.21298   .0323965    68.31   0.000     2.149466    2.276494
-------------------------------------------------------------------------------

. 
. ** Row 7: "Confidence that Social Security will Pay Benefits (Standardized)"
. reg logspread ssa_Sconfident                             , robust

Linear regression                               Number of obs     =      3,115
                                                F(1, 3113)        =       9.50
                                                Prob > F          =     0.0021
                                                R-squared         =     0.0030
                                                Root MSE          =     2.1012

--------------------------------------------------------------------------------
               |               Robust
     logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
ssa_Sconfident |  -.1157039   .0375408    -3.08   0.002    -.1893111   -.0420967
         _cons |    2.14096   .0376475    56.87   0.000     2.067144    2.214777
--------------------------------------------------------------------------------

. 
. ** Row 8: "Receives Annuity Income (Dummy)"
. reg logspread afin_annuity                               , robust

Linear regression                               Number of obs     =      4,420
                                                F(1, 4418)        =       0.22
                                                Prob > F          =     0.6372
                                                R-squared         =     0.0001
                                                Root MSE          =     2.1413

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
afin_annuity |  -.1107271   .2347936    -0.47   0.637    -.5710402    .3495861
       _cons |    2.24629   .0325181    69.08   0.000     2.182539    2.310042
------------------------------------------------------------------------------

. 
. ** Row 9: "Owns an IRA or Keogh (Dummy)"
. reg logspread afin_irakeogh                              , robust

Linear regression                               Number of obs     =      4,422
                                                F(1, 4420)        =     106.13
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0220
                                                Root MSE          =     2.1178

-------------------------------------------------------------------------------
              |               Robust
    logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
afin_irakeogh |  -.6629639   .0643532   -10.30   0.000    -.7891283   -.5367994
        _cons |   2.480789   .0412536    60.14   0.000     2.399912    2.561667
-------------------------------------------------------------------------------

. 
. ** Row 10: "Ability and Comfort with Retirement Planning (Standardized)"
. reg logspread aplan_index                                , robust

Linear regression                               Number of obs     =      3,474
                                                F(1, 3472)        =      60.83
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0182
                                                Root MSE          =     2.1165

------------------------------------------------------------------------------
             |               Robust
   logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
 aplan_index |  -.2885339    .036995    -7.80   0.000    -.3610682   -.2159997
       _cons |   2.221201   .0359153    61.85   0.000     2.150784    2.291618
------------------------------------------------------------------------------

. 
. 
. 
. 
.         
. // ***************************************************************************************
. //                      APX TABLE A10: Characteristics of People with Buy Values Exceeding Sell Values
. // ***************************************************************************************
. 
. ** Define a categorical variable for differences between the buy and sell value
. ** (It is more than we need here, but it will be used later on again)
. gen logsellbuydiff_cats = .
(4,596 missing values generated)

. replace logsellbuydiff_cats = 1 if logsellbuydiff <  -1 
(814 real changes made)

. replace logsellbuydiff_cats = 2 if logsellbuydiff >= -1 &  logsellbuydiff <0  
(447 real changes made)

. replace logsellbuydiff_cats = 3 if logsellbuydiff ==0
(450 real changes made)

. replace logsellbuydiff_cats = 4 if logsellbuydiff > 0   &  logsellbuydiff <=1
(877 real changes made)

. replace logsellbuydiff_cats = 5 if logsellbuydiff > 1   &  logsellbuydiff <.
(1,964 real changes made)

. 
. label def logsellbuydiff_cats 1 "logdif<-1" 2 "-1<=logdif<0" 3 "logdif=0" 4 "0<logdif<=1" 5 "1<logdif"

. label val logsellbuydiff_cats logsellbuydiff_cats 

. 
. 
. ** Summarize variables for the group with buy>sell
. sum agecat_* $demographics_balance cognix_pca admc_Sframe admc_Stime admc_Ssubset ssa_Sknowledge ///
>              ssa_Sliteracy ssa_Sconfident afin_annuity afin_irakeogh aplan_index ///
>              if basesample & (logsellbuydiff_cats==1 | logsellbuydiff_cats==2), sep(0)   

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
agecat_18_34 |      1,110    .2765766    .4475069          0          1
agecat_35_49 |      1,110          .3    .4584641          0          1
agecat_50_64 |      1,110    .3027027    .4596348          0          1
agecat_65_~s |      1,110    .1207207    .3259493          0          1
         age |      1,110    45.94414    15.56071         18         94
       agesq |      1,110    23.52782    15.17516       3.24      88.36
      female |      1,110    .6036036     .489369          0          1
     married |      1,110    .5522523    .4974864          0          1
     nhwhite |      1,110    .7207207    .4488473          0          1
     nhblack |      1,110     .090991    .2877259          0          1
     nhother |      1,110     .081982    .2744609          0          1
    hispanic |      1,110    .1063063    .3083682          0          1
  ed_dropout |      1,110    .0612613     .239917          0          1
  ed_hschool |      1,110    .2009009    .4008547          0          1
  ed_somecol |      1,110    .3972973    .4895591          0          1
  ed_college |      1,110     .209009    .4067841          0          1
  ed_graduat |      1,110    .1315315    .3381331          0          1
   hinc_lt25 |      1,110    .1972973     .398138          0          1
  hinc_25_50 |      1,110    .1720721    .3776132          0          1
  hinc_50_75 |      1,110    .1702703    .3760395          0          1
 hinc_75_100 |      1,110    .1351351     .342022          0          1
  hinc_ge100 |      1,110    .3252252    .4686701          0          1
     hhsiz_1 |      1,110     .218018    .4130858          0          1
     hhsiz_2 |      1,110    .3531532    .4781652          0          1
     hhsiz_3 |      1,110    .1954955    .3967604          0          1
    hhsiz_4p |      1,110    .2333333    .4231432          0          1
     anykids |      1,110    .3585586    .4797934          0          1
  cognix_pca |      1,110   -.1689746    1.005285  -3.649997   2.058814
 admc_Sframe |      1,110   -.0772797    .9976547  -2.578724   1.057953
  admc_Stime |      1,110   -.0171146    1.014666  -4.169115    1.45912
admc_Ssubset |      1,110   -.0389771    1.028033  -4.669462    .335249
ssa_Sknowl~e |      1,041   -.0972408    1.017545  -2.012349    1.55509
ssa_Sliter~y |      1,043   -.1038123    1.017635  -3.395885   1.349978
ssa_Sconfi~t |        769   -.0245887    1.001482   -1.10041   2.058695
afin_annuity |      1,090    .0165138    .1274989          0          1
afin_irake~h |      1,089    .3085399    .4621029          0          1
 aplan_index |        882   -.0331568    .9628906   -2.78208   2.093021

. 
. ** Summarize variables for the  group with buy<sell
. sum agecat_* $demographics_balance cognix_pca admc_Sframe admc_Stime admc_Ssubset ssa_Sknowledge ///
>              ssa_Sliteracy ssa_Sconfident afin_annuity afin_irakeogh aplan_index ///
>              if basesample & (logsellbuydiff_cats==4 | logsellbuydiff_cats==5), sep(0)  

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
agecat_18_34 |      2,542    .2053501    .4040367          0          1
agecat_35_49 |      2,542    .2867821    .4523479          0          1
agecat_50_64 |      2,542    .3276947    .4694652          0          1
agecat_65_~s |      2,542    .1801731    .3844072          0          1
         age |      2,542    49.52636    15.40192         18        106
       agesq |      2,542    26.89986    15.76267       3.24     112.36
      female |      2,542    .5735641     .494656          0          1
     married |      2,542    .6030685    .4893578          0          1
     nhwhite |      2,542    .7627852    .4254587          0          1
     nhblack |      2,542    .0798584    .2711272          0          1
     nhother |      2,542    .0767113    .2661851          0          1
    hispanic |      2,542    .0806452     .272343          0          1
  ed_dropout |      2,542    .0527144    .2235067          0          1
  ed_hschool |      2,542    .1982691    .3987744          0          1
  ed_somecol |      2,542    .3945712    .4888546          0          1
  ed_college |      2,542     .211251    .4082764          0          1
  ed_graduat |      2,542    .1431943    .3503398          0          1
   hinc_lt25 |      2,542    .1612903    .3678709          0          1
  hinc_25_50 |      2,542    .1797797    .3840794          0          1
  hinc_50_75 |      2,542    .1656176    .3718102          0          1
 hinc_75_100 |      2,542    .1219512    .3272938          0          1
  hinc_ge100 |      2,542    .3713611    .4832638          0          1
     hhsiz_1 |      2,542    .1986625    .3990719          0          1
     hhsiz_2 |      2,542     .402439     .490486          0          1
     hhsiz_3 |      2,542    .1695515    .3753122          0          1
    hhsiz_4p |      2,542     .229347    .4204955          0          1
     anykids |      2,542    .3143194    .4643356          0          1
  cognix_pca |      2,542   -.0014129    .9762925  -3.721495   2.184767
 admc_Sframe |      2,536   -.0139774    1.002589  -2.578724   1.057953
  admc_Stime |      2,536   -.0005687    1.007917  -4.973148    1.45912
admc_Ssubset |      2,536   -.0002403    1.004411  -4.669462    .335249
ssa_Sknowl~e |      2,419    .0455281    .9913826  -2.012349    1.55509
ssa_Sliter~y |      2,426    .0510278    .9930176  -3.989118   1.349978
ssa_Sconfi~t |      1,760    .0038948    1.001086   -1.10041   2.058695
afin_annuity |      2,496     .021234    .1441923          0          1
afin_irake~h |      2,498    .3742994    .4840384          0          1
 aplan_index |      1,911    .0108351    1.017733   -2.78208   2.093021

. 
.   
. ** Calculate the significance of the differences using t-tests
. gen tmp=logsellbuydiff_cats
(44 missing values generated)

. recode tmp 1 2 = 1 3=. 4 5=0    /* dummy for being in buy>sell group as opposed to sell>buy */
(tmp: 3738 changes made)

. foreach x of varlist agecat_* $demographics_balance cognix_pca admc_Sframe admc_Stime admc_Ssubset ///
>                      ssa_Sknowledge ssa_Sliteracy ssa_Sconfident afin_annuity afin_irakeogh aplan_index {
  2.   qui ttest `x', by(tmp) unequal
  3.   di "P-value for t-test of difference is "  %8.4f r(p) "  for `x' "
  4. }   
P-value for t-test of difference is   0.0000  for agecat_18_34 
P-value for t-test of difference is   0.3970  for agecat_35_49 
P-value for t-test of difference is   0.1429  for agecat_50_64 
P-value for t-test of difference is   0.0000  for agecat_65_plus 
P-value for t-test of difference is   0.0000  for age 
P-value for t-test of difference is   0.0000  for agesq 
P-value for t-test of difference is   0.2120  for female 
P-value for t-test of difference is   0.0019  for married 
P-value for t-test of difference is   0.0012  for nhwhite 
P-value for t-test of difference is   0.3006  for nhblack 
P-value for t-test of difference is   0.2606  for nhother 
P-value for t-test of difference is   0.0058  for hispanic 
P-value for t-test of difference is   0.3420  for ed_dropout 
P-value for t-test of difference is   0.8432  for ed_hschool 
P-value for t-test of difference is   0.9525  for ed_somecol 
P-value for t-test of difference is   0.8848  for ed_college 
P-value for t-test of difference is   0.5278  for ed_graduat 
P-value for t-test of difference is   0.0065  for hinc_lt25 
P-value for t-test of difference is   0.5216  for hinc_25_50 
P-value for t-test of difference is   0.6847  for hinc_50_75 
P-value for t-test of difference is   0.2965  for hinc_75_100 
P-value for t-test of difference is   0.0324  for hinc_ge100 
P-value for t-test of difference is   0.1393  for hhsiz_1 
P-value for t-test of difference is   0.0128  for hhsiz_2 
P-value for t-test of difference is   0.4746  for hhsiz_3 
P-value for t-test of difference is   0.4536  for hhsiz_4p 
P-value for t-test of difference is   0.0254  for anykids 
P-value for t-test of difference is   0.0000  for cognix_pca 
P-value for t-test of difference is   0.0725  for admc_Sframe 
P-value for t-test of difference is   0.6697  for admc_Stime 
P-value for t-test of difference is   0.5714  for admc_Ssubset 
P-value for t-test of difference is   0.0000  for ssa_Sknowledge 
P-value for t-test of difference is   0.0000  for ssa_Sliteracy 
P-value for t-test of difference is   0.1704  for ssa_Sconfident 
P-value for t-test of difference is   0.2968  for afin_annuity 
P-value for t-test of difference is   0.0016  for afin_irakeogh 
P-value for t-test of difference is   0.1687  for aplan_index 

. 
. 
. ** test for joint significance in a regression, to report in the last row of the table
. reg tmp $demographics cognix_pca admc__Sframe* admc__Stime* admc__Ssubset* ssa__Sknowledge* ///
>         ssa__Sliteracy* ssa__Sconfident* afin__annuity* afin__irakeogh* aplan__index* if basesample, robust
note: admc__Stime_m omitted because of collinearity
note: admc__Ssubset_m omitted because of collinearity

Linear regression                               Number of obs     =      3,652
                                                F(36, 3615)       =       8.53
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0231
                                                Root MSE          =     .45693

-----------------------------------------------------------------------------------
                  |               Robust
              tmp |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
------------------+----------------------------------------------------------------
              age |  -.0049314   .0028903    -1.71   0.088    -.0105981    .0007354
            agesq |   .0019233    .002878     0.67   0.504    -.0037193    .0075659
           female |  -.0056979   .0163161    -0.35   0.727    -.0376876    .0262918
          married |  -.0082895   .0184927    -0.45   0.654    -.0445468    .0279677
          nhblack |  -.0201405   .0307818    -0.65   0.513    -.0804919     .040211
          nhother |  -.0012305   .0293326    -0.04   0.967    -.0587406    .0562795
         hispanic |   .0140176   .0307666     0.46   0.649     -.046304    .0743393
       ed_dropout |  -.0226304   .0367759    -0.62   0.538     -.094734    .0494733
       ed_hschool |  -.0172816   .0213953    -0.81   0.419    -.0592295    .0246664
       ed_college |   .0238257    .021958     1.09   0.278    -.0192255     .066877
       ed_graduat |   .0200104   .0250567     0.80   0.425    -.0291162    .0691371
       hinc_25_50 |  -.0270463   .0265568    -1.02   0.309    -.0791141    .0250216
       hinc_50_75 |  -.0050056   .0278794    -0.18   0.858    -.0596666    .0496554
      hinc_75_100 |   .0139122   .0309092     0.45   0.653    -.0466891    .0745134
       hinc_ge100 |  -.0150405    .026411    -0.57   0.569    -.0668225    .0367415
          hhsiz_2 |  -.0304967   .0229418    -1.33   0.184    -.0754769    .0144835
          hhsiz_3 |  -.0181379   .0308395    -0.59   0.556    -.0786025    .0423267
         hhsiz_4p |   -.065205   .0348215    -1.87   0.061    -.1334767    .0030666
          anykids |   .0346652   .0266355     1.30   0.193    -.0175569    .0868872
       cognix_pca |  -.0286249   .0110725    -2.59   0.010    -.0503338    -.006916
   admc__Sframe_m |  -.2814591   .0330846    -8.51   0.000    -.3463256   -.2165927
     admc__Sframe |   .0035999   .0083576     0.43   0.667    -.0127862     .019986
    admc__Stime_m |          0  (omitted)
      admc__Stime |   .0005128   .0078823     0.07   0.948    -.0149413     .015967
  admc__Ssubset_m |          0  (omitted)
    admc__Ssubset |    .000147   .0078957     0.02   0.985    -.0153336    .0156275
ssa__Sknowledge_m |  -.0372157   .0709522    -0.52   0.600    -.1763261    .1018947
  ssa__Sknowledge |  -.0154299   .0090237    -1.71   0.087     -.033122    .0022622
 ssa__Sliteracy_m |   .0018417    .074532     0.02   0.980    -.1442872    .1479707
   ssa__Sliteracy |  -.0045691   .0093195    -0.49   0.624     -.022841    .0137029
ssa__Sconfident_m |   .0464909   .0245412     1.89   0.058     -.001625    .0946069
  ssa__Sconfident |   .0135866    .009803     1.39   0.166    -.0056334    .0328066
  afin__annuity_m |  -.3657116   .1362799    -2.68   0.007    -.6329048   -.0985185
    afin__annuity |   .0092156   .0534738     0.17   0.863    -.0956262    .1140574
 afin__irakeogh_m |   .3866283   .1426832     2.71   0.007     .1068806    .6663759
   afin__irakeogh |  -.0177588   .0185746    -0.96   0.339    -.0541765    .0186589
   aplan__index_m |   .0017667   .0265927     0.07   0.947    -.0503715    .0539049
     aplan__index |   .0050018   .0095375     0.52   0.600    -.0136975    .0237011
            _cons |   .5186932   .0777943     6.67   0.000     .3661681    .6712183
-----------------------------------------------------------------------------------

. 
. drop tmp

. 
. 
.     
. // ***************************************************************************************
. //       TEXT CLAIMS: Results mentioned in the text but that don't appear in tables
. // ***************************************************************************************
. 
. 
. ** Text Claim 01 - Section 2.4 (Footnote #9)    
. ** -----------------------------------------
. ** "Respondents took on average about 30% longer to read and process the vignettes 
. ** of the complexity treatment than the control vignette (“no added complexity”), 
. ** and the text of vignettes of the complexity treatment required a reading 
. ** comprehension 0.9 grade levels higher, according to the Flesch-Kincaid scale."
. **
. ** Note: the comprehension grade levels were found by importing the text of the 
. **       vignette into MS Word, and using the reading comprehension tool in
. **       MS Word
. 
. 
. tab complexity if basesample, sum(time_ad_intro)

  indicates |
the intro R |
   received |
    for the |
     advice |      Summary of ad_intro time_
  questions |        Mean   Std. Dev.       Freq.
------------+------------------------------------
  1 Narrow  |   42.040143   35.151586       1,395
  2 Wide sp |   47.670008   37.583557       1,297
  3 Adding  |   62.112623   44.251125       1,323
------------+------------------------------------
      Total |   50.472976   40.024995       4,015

. reg time_ad_intro complexity_2 complexity_3 if basesample, robust 

Linear regression                               Number of obs     =      4,015
                                                F(2, 4012)        =      86.23
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0449
                                                Root MSE          =     39.126

------------------------------------------------------------------------------
             |               Robust
time_ad_in~o |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
complexity_2 |   5.629864   1.405286     4.01   0.000     2.874722    8.385006
complexity_3 |   20.07248   1.538137    13.05   0.000     17.05688    23.08808
       _cons |   42.04014   .9411621    44.67   0.000     40.19494    43.88534
------------------------------------------------------------------------------

. di "Percent increase in reading time for complexity treatments:      "  50*(_b[complexity_2]+_b[complexity_3])/_b[_cons]
Percent increase in reading time for complexity treatments:      30.568811

. di "Percent increase in reading time for complexity: wide age range: " 100*_b[complexity_2]/_b[_cons]
Percent increase in reading time for complexity: wide age range: 13.391639

. di "Percent increase in reading time for complexity: extra info:     " 100*_b[complexity_3]/_b[_cons]
Percent increase in reading time for complexity: extra info:     47.745982

. 
. 
. 
. 
. ** Text Claim 02 - Section 2.4 
. ** ---------------------------
. ** "These factual questions are two multiple choice questions about the financial 
. ** advisor’s explanation of the benefits and drawbacks under each scenario (spending
. ** down slowly or quickly). Of the respondents who are posed these two factual questions, 
. ** 63% answer both correctly, 27% answer one correctly, and 10% answer neither correctly."
. 
. ** What fraction of people correctly answered the follow-up questions to the consequential intervention?
. gen correct_followup1=(test_question1==3) if consequence & basesample
(2,534 missing values generated)

. gen correct_followup2=(test_question2==3) if consequence & basesample
(2,534 missing values generated)

. gen correct_followup_total = correct_followup1+correct_followup2
(2,534 missing values generated)

. 
. label var correct_followup_total "Total Questions Correct (0-2)"

. 
. ** As expected, the number of correct answers varies strongly by cognition
. ** Last column shows the percentages answering both correctly and exactly one correctly.
. tab correct_followup_total cognix_xtile4 if consequence & basesample, col nof

     Total |
 Questions |
   Correct |    Quartiles of the PCA cognition index
     (0-2) |         1          2          3          4 |     Total
-----------+--------------------------------------------+----------
         0 |     23.87       8.80       5.85       1.31 |      9.84 
         1 |     40.24      32.31      22.98      12.13 |     26.77 
         2 |     35.90      58.89      71.17      86.57 |     63.39 
-----------+--------------------------------------------+----------
     Total |    100.00     100.00     100.00     100.00 |    100.00 


. 
. 
.         
.         
. ** Text Claim 03 - Section 3.2  
. ** ---------------------------
. ** "Respondents advise our hypothetical vignette individuals to buy an annuity 
. ** that pays $100 per month for a median price of $4,750 (s.e.: $180) but advise 
. ** them to sell this annuity for a median price of $16,250 (s.e.: $543)."
.  
. ** summarize to get the medians and run median regressions to get standard errors 
. sum  midbuy  if ~sell_first & basesample, d

                    Buy price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          250            250
10%          250            250       Obs               2,051
25%         1250            250       Sum of Wgt.       2,051

50%         4750                      Mean           21059.61
                        Largest       Std. Dev.      90082.72
75%        13750         750000
90%        23750         750000       Variance       8.11e+09
95%        42500         750000       Skewness       7.430982
99%       750000         750000       Kurtosis       58.34003

. qreg midbuy  if ~sell_first & basesample
Iteration  1:  WLS sum of weighted deviations =   25078320

Iteration  1: sum of abs. weighted deviations =   24310250
Iteration  2: sum of abs. weighted deviations =   19833125

Median regression                                   Number of obs =      2,051
  Raw sum of deviations 1.98e+07 (about 4750)
  Min sum of deviations 1.98e+07                    Pseudo R2     =     0.0000

------------------------------------------------------------------------------
      midbuy |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |       4750   180.4748    26.32   0.000     4396.067    5103.933
------------------------------------------------------------------------------

. 
. sum  midsell if  sell_first & basesample, d

                    Sell price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%         1250            250
10%         2750            250       Obs               2,009
25%         7375            250       Sum of Wgt.       2,009

50%        16250                      Mean           57658.54
                        Largest       Std. Dev.      143146.2
75%        31250         750000
90%        95000         750000       Variance       2.05e+10
95%       350000         750000       Skewness       3.856013
99%       750000         750000       Kurtosis       17.24059

. qreg midsell if  sell_first & basesample
Iteration  1:  WLS sum of weighted deviations =   66086509

Iteration  1: sum of abs. weighted deviations =   65428250
Iteration  2: sum of abs. weighted deviations =   50672625

Median regression                                   Number of obs =      2,009
  Raw sum of deviations 5.07e+07 (about 16250)
  Min sum of deviations 5.07e+07                    Pseudo R2     =     0.0000

------------------------------------------------------------------------------
     midsell |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |      16250   543.2946    29.91   0.000     15184.52    17315.48
------------------------------------------------------------------------------

.    
.    
.    
.    
. ** Text Claim 04 - Section 3.2  
. ** ---------------------------
. ** "This represents a statistically significant difference (two-sample 
. ** Wilcoxon-Mann-Whitney rank-sum test z-statistic=25.8, p-value<0.001)."
. **
. ** These are from different (and independent) samples because sell_first randomized
. ** across respondents. Hence to an exact  Wilcoxon-Mann-Whitney ranksum test.
. ** To implement this, the data from the different groups needs to be in the 
. ** same variable. Create this variable first:
. 
. gen     valuation=.
(4,596 missing values generated)

. replace valuation=midbuy  if ~sell_first & basesample
(2,051 real changes made)

. replace valuation=midsell if  sell_first & basesample
(2,009 real changes made)

. ranksumex valuation, by(sell_first)

Two-sample Wilcoxon rank-sum (Mann-Whitney) test

  sell_first |      obs    rank sum    expected
-------------+---------------------------------
          No |     2051     3200397   4164555.5
         Yes |     2009     5043433   4079274.5
-------------+---------------------------------
    combined |     4060     8243830     8243830

unadjusted variance   1.394e+09
adjustment for ties  -2036981.9
                     ----------
adjusted variance     1.392e+09

Ho: valuat~n(sell_f~t==No) = valuat~n(sell_f~t==Yes)
             z = -25.838
    Prob > |z| =   0.0000

. drop valuation

. 
. 
.    
. ** Text Claim 05 - Section 3.2  
. ** ---------------------------
. ** "Only about 10 percent of respondents have a buy value that is equal to their 
. ** sell value, and only 40 percent have a buy and sell value that are within one
. ** log unit (i.e., within a factor of 2.72) of each other."
. 
. tab logsellbuydiff_cats if basesample

logsellbuydi |
     ff_cats |      Freq.     Percent        Cum.
-------------+-----------------------------------
   logdif<-1 |        709       17.46       17.46
-1<=logdif<0 |        401        9.88       27.34
    logdif=0 |        408       10.05       37.39
 0<logdif<=1 |        796       19.61       57.00
    1<logdif |      1,746       43.00      100.00
-------------+-----------------------------------
       Total |      4,060      100.00

. 
. 
. 
. ** Text Claim 06 - Section 3.2 (answered by same output as text claim 05)       
. ** ----------------------------------------------------------------------
. ** "Second, the distribution is not symmetric around zero: 63% have sell valuations 
. ** that strictly exceed their buy valuations, whereas buy valuations strictly exceed
. ** sell valuations for about 27% of respondents."
. 
. tab logsellbuydiff_cats if basesample

logsellbuydi |
     ff_cats |      Freq.     Percent        Cum.
-------------+-----------------------------------
   logdif<-1 |        709       17.46       17.46
-1<=logdif<0 |        401        9.88       27.34
    logdif=0 |        408       10.05       37.39
 0<logdif<=1 |        796       19.61       57.00
    1<logdif |      1,746       43.00      100.00
-------------+-----------------------------------
       Total |      4,060      100.00

. 
. 
.    
. ** Text Claim 07 - Section 3.2  
. ** ---------------------------
. ** "A further similarity is that we also find that the log buy and the log sell 
. ** valuations are negatively correlated (correlation coefficient: -0.11, p-value<0.001)."
. 
. pwcorr logbuyprice logsellprice if basesample, sig

             | logbuy~e logsel~e
-------------+------------------
 logbuyprice |   1.0000 
             |
             |
logsellprice |  -0.1066   1.0000 
             |   0.0000
             |

. 
. 
. 
. 
. ** Text Claim 08 - Section 3.3 (Footnote #15)   
. ** ------------------------------------------
. ** "We do not control for the order in which the two blocks of consequence message
. ** treatment were shown because this variable is available for only half the sample. 
. ** Within the half of the sample for which this order was randomized, the order 
. ** has no significant effect on the spread (p-value: 0.758)."
. 
. ** There is no effect of the order of the test questions used in the consequence treatment on the spread
. ** Look at P>(t) value for quick_first variable
. reg logspread    any_complexity quick_first consequence cognix_pca $exp_controls $demographics if basesample & consequence, robust
note: consequence omitted because of collinearity

Linear regression                               Number of obs     =      2,062
                                                F(32, 2029)       =      11.97
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1609
                                                Root MSE          =     1.9205

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .0711834   .0895624     0.79   0.427    -.1044605    .2468273
    quick_first |  -.0261306   .0847788    -0.31   0.758    -.1923932    .1401319
    consequence |          0  (omitted)
     cognix_pca |  -.7836279   .0600138   -13.06   0.000    -.9013231   -.6659328
     sell_first |   .2268299   .0849633     2.67   0.008     .0602055    .3934543
ls_startvalue_2 |   .0959821   .1038025     0.92   0.355    -.1075886    .2995527
ls_startvalue_3 |   .0963233   .1040448     0.93   0.355    -.1077225    .3003691
       ls_first |   .0112557    .085086     0.13   0.895    -.1556093    .1781207
   ss_benefit_2 |   .1495711   .1237436     1.21   0.227    -.0931068    .3922489
   ss_benefit_3 |   .0785101   .1156371     0.68   0.497    -.1482697      .30529
   ss_benefit_4 |   .1846492   .1201208     1.54   0.124    -.0509238    .4202221
vignette_name_2 |   .1314423   .1194154     1.10   0.271    -.1027473    .3656319
vignette_name_3 |  -.0282263    .121594    -0.23   0.816    -.2666884    .2102357
vignette_name_4 |  -.0428672   .1174952    -0.36   0.715     -.273291    .1875565
            age |   .0287136   .0177487     1.62   0.106    -.0060939    .0635211
          agesq |   -.017641   .0178478    -0.99   0.323    -.0526429     .017361
         female |   .0836664   .0908964     0.92   0.357    -.0945936    .2619263
        married |   .1382138   .1007074     1.37   0.170    -.0592868    .3357144
        nhblack |  -.0582776    .188494    -0.31   0.757    -.4279396    .3113843
        nhother |  -.0575517    .164893    -0.35   0.727    -.3809289    .2658254
       hispanic |   .2389864   .1793831     1.33   0.183    -.1128079    .5907807
     ed_dropout |   -.247275   .2481717    -1.00   0.319    -.7339729    .2394229
     ed_hschool |  -.2054507   .1268924    -1.62   0.106    -.4543037    .0434023
     ed_college |  -.0088448   .1173567    -0.08   0.940     -.238997    .2213073
     ed_graduat |   .0439282   .1310739     0.34   0.738    -.2131253    .3009817
     hinc_25_50 |  -.1643907   .1596888    -1.03   0.303    -.4775618    .1487803
     hinc_50_75 |  -.4014872   .1593977    -2.52   0.012    -.7140875   -.0888869
    hinc_75_100 |  -.1408782   .1780527    -0.79   0.429    -.4900633    .2083069
     hinc_ge100 |  -.4267036   .1468292    -2.91   0.004    -.7146553    -.138752
        hhsiz_2 |   .0623496   .1278872     0.49   0.626    -.1884543    .3131534
        hhsiz_3 |   .2256867   .1822455     1.24   0.216    -.1317212    .5830945
       hhsiz_4p |   .4121571   .1991476     2.07   0.039      .021602    .8027121
        anykids |  -.2603884   .1468086    -1.77   0.076    -.5482997     .027523
          _cons |   .9763126   .4718641     2.07   0.039     .0509239    1.901701
---------------------------------------------------------------------------------

. 
. 
. 
.  
. ** Text Claim 09 - Section 3.3  
. ** ---------------------------
. ** "While the estimates seem to indicate that the complexity treatment primarily 
. ** operates on the buy price, and hence it reduces the average of the log sell and 
. ** buy price, this is not a valid interpretation as we cannot reject that increase
. ** in the sell price and the decrease in the buy price are the same in absolute 
. ** value (p-value: 0.302)."
.  
. ** The effect of the treatments on the mean annuity valuation
. ** Look at P>(t) value for any_complexity variable
. reg meanlogprice any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =       7.84
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0580
                                                Root MSE          =     1.2717

---------------------------------------------------------------------------------
                |               Robust
   meanlogprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |  -.0435194    .042162    -1.03   0.302    -.1261804    .0391415
    consequence |   .0720687    .040139     1.80   0.073    -.0066258    .1507633
     cognix_pca |  -.0450744   .0273086    -1.65   0.099    -.0986143    .0084656
     sell_first |   .3672457   .0402471     9.12   0.000     .2883391    .4461523
ls_startvalue_2 |   .2374405   .0484064     4.91   0.000     .1425372    .3323438
ls_startvalue_3 |   .4803635   .0494642     9.71   0.000     .3833863    .5773407
       ls_first |  -.0543333   .0400979    -1.36   0.175    -.1329473    .0242808
   ss_benefit_2 |  -.2241979   .0565026    -3.97   0.000    -.3349744   -.1134215
   ss_benefit_3 |  -.1997549   .0559645    -3.57   0.000    -.3094762   -.0900336
   ss_benefit_4 |  -.2359684   .0579316    -4.07   0.000    -.3495463   -.1223905
vignette_name_2 |  -.0634025   .0550835    -1.15   0.250    -.1713967    .0445917
vignette_name_3 |   .0084243   .0560327     0.15   0.880    -.1014308    .1182793
vignette_name_4 |   .0325216   .0556981     0.58   0.559    -.0766774    .1417206
            age |  -.0166653    .007272    -2.29   0.022    -.0309224   -.0024081
          agesq |   .0145784   .0070847     2.06   0.040     .0006886    .0284683
         female |  -.1175297   .0429174    -2.74   0.006    -.2016716   -.0333879
        married |  -.0553122   .0506293    -1.09   0.275    -.1545736    .0439493
        nhblack |  -.1015667   .0911993    -1.11   0.265    -.2803677    .0772343
        nhother |   -.071495   .0777334    -0.92   0.358    -.2238955    .0809055
       hispanic |  -.0955904   .0830753    -1.15   0.250    -.2584639    .0672831
     ed_dropout |   .1368372   .1082717     1.26   0.206    -.0754352    .3491096
     ed_hschool |   .0756271   .0615101     1.23   0.219    -.0449667    .1962208
     ed_college |    .059729    .053902     1.11   0.268    -.0459487    .1654067
     ed_graduat |    .228132   .0593661     3.84   0.000     .1117415    .3445225
     hinc_25_50 |  -.0447785   .0745103    -0.60   0.548      -.19086    .1013029
     hinc_50_75 |  -.0537256   .0762761    -0.70   0.481    -.2032688    .0958177
    hinc_75_100 |  -.0573641    .082673    -0.69   0.488     -.219449    .1047208
     hinc_ge100 |  -.0695452   .0689813    -1.01   0.313    -.2047868    .0656963
        hhsiz_2 |    .017321   .0596804     0.29   0.772    -.0996856    .1343276
        hhsiz_3 |   .1136685   .0794728     1.43   0.153     -.042142    .2694791
       hhsiz_4p |  -.0145206   .0944646    -0.15   0.878    -.1997235    .1706823
        anykids |  -.0609772   .0732333    -0.83   0.405    -.2045551    .0826007
          _cons |   9.477364     .20641    45.92   0.000     9.072687    9.882042
---------------------------------------------------------------------------------

. 
. 
. 
. 
. ** Text Claim 10 - Section 3.3 (answered by same output as text claim 09)
. ** ----------------------------------------------------------------------
. ** "In fact, it marginally significantly increases the average of the log buy and 
. ** sell price (p-value 0.073), suggesting that the consequence message not only 
. ** increases the rationality of the annuity valuations but also raises the levels."
. 
. ** Effect of the treatments on the mean annuity valuation
. reg meanlogprice any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =       7.84
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0580
                                                Root MSE          =     1.2717

---------------------------------------------------------------------------------
                |               Robust
   meanlogprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |  -.0435194    .042162    -1.03   0.302    -.1261804    .0391415
    consequence |   .0720687    .040139     1.80   0.073    -.0066258    .1507633
     cognix_pca |  -.0450744   .0273086    -1.65   0.099    -.0986143    .0084656
     sell_first |   .3672457   .0402471     9.12   0.000     .2883391    .4461523
ls_startvalue_2 |   .2374405   .0484064     4.91   0.000     .1425372    .3323438
ls_startvalue_3 |   .4803635   .0494642     9.71   0.000     .3833863    .5773407
       ls_first |  -.0543333   .0400979    -1.36   0.175    -.1329473    .0242808
   ss_benefit_2 |  -.2241979   .0565026    -3.97   0.000    -.3349744   -.1134215
   ss_benefit_3 |  -.1997549   .0559645    -3.57   0.000    -.3094762   -.0900336
   ss_benefit_4 |  -.2359684   .0579316    -4.07   0.000    -.3495463   -.1223905
vignette_name_2 |  -.0634025   .0550835    -1.15   0.250    -.1713967    .0445917
vignette_name_3 |   .0084243   .0560327     0.15   0.880    -.1014308    .1182793
vignette_name_4 |   .0325216   .0556981     0.58   0.559    -.0766774    .1417206
            age |  -.0166653    .007272    -2.29   0.022    -.0309224   -.0024081
          agesq |   .0145784   .0070847     2.06   0.040     .0006886    .0284683
         female |  -.1175297   .0429174    -2.74   0.006    -.2016716   -.0333879
        married |  -.0553122   .0506293    -1.09   0.275    -.1545736    .0439493
        nhblack |  -.1015667   .0911993    -1.11   0.265    -.2803677    .0772343
        nhother |   -.071495   .0777334    -0.92   0.358    -.2238955    .0809055
       hispanic |  -.0955904   .0830753    -1.15   0.250    -.2584639    .0672831
     ed_dropout |   .1368372   .1082717     1.26   0.206    -.0754352    .3491096
     ed_hschool |   .0756271   .0615101     1.23   0.219    -.0449667    .1962208
     ed_college |    .059729    .053902     1.11   0.268    -.0459487    .1654067
     ed_graduat |    .228132   .0593661     3.84   0.000     .1117415    .3445225
     hinc_25_50 |  -.0447785   .0745103    -0.60   0.548      -.19086    .1013029
     hinc_50_75 |  -.0537256   .0762761    -0.70   0.481    -.2032688    .0958177
    hinc_75_100 |  -.0573641    .082673    -0.69   0.488     -.219449    .1047208
     hinc_ge100 |  -.0695452   .0689813    -1.01   0.313    -.2047868    .0656963
        hhsiz_2 |    .017321   .0596804     0.29   0.772    -.0996856    .1343276
        hhsiz_3 |   .1136685   .0794728     1.43   0.153     -.042142    .2694791
       hhsiz_4p |  -.0145206   .0944646    -0.15   0.878    -.1997235    .1706823
        anykids |  -.0609772   .0732333    -0.83   0.405    -.2045551    .0826007
          _cons |   9.477364     .20641    45.92   0.000     9.072687    9.882042
---------------------------------------------------------------------------------

. 
. 
. 
. 
. 
. ** Text Claim 11 - Section 3.3 (Footnote #17)   
. ** ------------------------------------------
. ** "One might expect that people with an initially higher Social Security benefit 
. ** place a lower value on a $100 change in Social Security benefits, since they are 
. ** already more highly annuitized. To test this, we run an alternative specification 
. ** in which the baseline Social Security benefit amount is included as a linear control
. ** instead of as a set of dummy variables. Both the buy and sell value decline in the 
. ** baseline amount of Social Security benefits. The effect is not significant for the sell
. ** value (p-value 0.145), but there is a significant 2.5% decline in the buy value for 
. ** each additional $100 in baseline Social Security benefits."
. 
. ** Effect of ss_benefit100dollar on the log spread
. reg logspread    any_complexity consequence cognix_pca $exp_controls_linben $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(30, 4029)       =      25.64
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1565
                                                Root MSE          =     1.9602

-------------------------------------------------------------------------------------
                    |               Robust
          logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------------+----------------------------------------------------------------
     any_complexity |   .1318834    .064818     2.03   0.042     .0048043    .2589625
        consequence |  -.1415688   .0616189    -2.30   0.022     -.262376   -.0207616
         cognix_pca |  -.7881427   .0427403   -18.44   0.000    -.8719373   -.7043482
         sell_first |   .1653911   .0616457     2.68   0.007     .0445315    .2862508
    ls_startvalue_2 |   .0654172     .07609     0.86   0.390    -.0837613    .2145957
    ls_startvalue_3 |   -.002396    .074954    -0.03   0.975    -.1493473    .1445552
           ls_first |   .0291818    .061615     0.47   0.636    -.0916177    .1499813
ss_benefit100dollar |   .0111242   .0068782     1.62   0.106    -.0023608    .0246093
    vignette_name_2 |   .1134852   .0855872     1.33   0.185     -.054313    .2812835
    vignette_name_3 |   .0863004   .0877694     0.98   0.326    -.0857763     .258377
    vignette_name_4 |      -.011   .0850656    -0.13   0.897    -.1777756    .1557755
                age |   .0244363   .0129659     1.88   0.060    -.0009839    .0498566
              agesq |   -.014393    .012886    -1.12   0.264    -.0396568    .0108707
             female |    .085561   .0663041     1.29   0.197    -.0444317    .2155537
            married |   .0965994   .0759838     1.27   0.204    -.0523709    .2455697
            nhblack |   .0293375   .1419664     0.21   0.836    -.2489952    .3076701
            nhother |   .0515866   .1213008     0.43   0.671      -.18623    .2894033
           hispanic |   .0838283   .1251323     0.67   0.503    -.1615001    .3291568
         ed_dropout |  -.0564824   .1778471    -0.32   0.751    -.4051611    .2921964
         ed_hschool |   .0353872   .0930366     0.38   0.704    -.1470159    .2177904
         ed_college |     .00867   .0838267     0.10   0.918    -.1556767    .1730167
         ed_graduat |   .0754924   .0952486     0.79   0.428    -.1112474    .2622323
         hinc_25_50 |   .1039722   .1167363     0.89   0.373    -.1248955      .33284
         hinc_50_75 |  -.1611328    .116037    -1.39   0.165    -.3886295    .0663639
        hinc_75_100 |  -.0532261   .1299754    -0.41   0.682    -.3080497    .2015975
         hinc_ge100 |  -.2528961   .1099464    -2.30   0.021    -.4684518   -.0373403
            hhsiz_2 |  -.0241711   .0953813    -0.25   0.800    -.2111712    .1628289
            hhsiz_3 |     .14711   .1306062     1.13   0.260    -.1089503    .4031704
           hhsiz_4p |   .1796156   .1451459     1.24   0.216    -.1049505    .4641818
            anykids |  -.1758292   .1055079    -1.67   0.096     -.382683    .0310246
              _cons |   1.038528   .3517357     2.95   0.003     .3489316    1.728125
-------------------------------------------------------------------------------------

. 
. ** Effect of ss_benefit100dollar on the log sell price
. reg logsellprice any_complexity consequence cognix_pca $exp_controls_linben $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(30, 4029)       =       4.87
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0347
                                                Root MSE          =     1.7381

-------------------------------------------------------------------------------------
                    |               Robust
       logsellprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------------+----------------------------------------------------------------
     any_complexity |   .0501334   .0574957     0.87   0.383      -.06259    .1628569
        consequence |   .0101935   .0547139     0.19   0.852     -.097076    .1174629
         cognix_pca |  -.1890194    .038129    -4.96   0.000    -.2637732   -.1142655
         sell_first |  -.0412564   .0545006    -0.76   0.449    -.1481077    .0655949
    ls_startvalue_2 |   .2389912    .066578     3.59   0.000     .1084616    .3695209
    ls_startvalue_3 |   .4846692   .0675329     7.18   0.000     .3522674     .617071
           ls_first |  -.0433498   .0547669    -0.79   0.429    -.1507233    .0640237
ss_benefit100dollar |  -.0091421   .0062765    -1.46   0.145    -.0214475    .0031633
    vignette_name_2 |  -.0280677   .0761231    -0.37   0.712    -.1773111    .1211758
    vignette_name_3 |  -.0949528   .0763623    -1.24   0.214    -.2446651    .0547595
    vignette_name_4 |  -.0817567   .0760793    -1.07   0.283    -.2309141    .0674008
                age |   .0015188   .0108285     0.14   0.888    -.0197111    .0227487
              agesq |   .0055583   .0103551     0.54   0.591    -.0147434    .0258599
             female |  -.0752539   .0584331    -1.29   0.198    -.1898152    .0393074
            married |  -.0071657   .0694355    -0.10   0.918    -.1432976    .1289662
            nhblack |   -.092135   .1341388    -0.69   0.492    -.3551212    .1708511
            nhother |  -.0583685   .1073573    -0.54   0.587    -.2688482    .1521111
           hispanic |  -.0976988   .1215931    -0.80   0.422    -.3360885    .1406909
         ed_dropout |   .1412997    .161061     0.88   0.380     -.174469    .4570684
         ed_hschool |   .1048103   .0854601     1.23   0.220    -.0627387    .2723594
         ed_college |   .0191268   .0730595     0.26   0.793    -.1241103    .1623639
         ed_graduat |   .2261839    .076512     2.96   0.003      .076178    .3761898
         hinc_25_50 |   .0465371    .108277     0.43   0.667    -.1657457    .2588199
         hinc_50_75 |  -.0688931   .1075312    -0.64   0.522    -.2797137    .1419275
        hinc_75_100 |  -.1031729   .1188747    -0.87   0.385    -.3362331    .1298873
         hinc_ge100 |  -.0988502   .1002722    -0.99   0.324     -.295439    .0977387
            hhsiz_2 |   .0432722   .0836213     0.52   0.605    -.1206717    .2072161
            hhsiz_3 |   .2518931   .1134611     2.22   0.026     .0294467    .4743396
           hhsiz_4p |   .1768586   .1346644     1.31   0.189     -.087158    .4408753
            anykids |   -.236776   .1009518    -2.35   0.019    -.4346974   -.0388547
              _cons |   9.438767   .3143288    30.03   0.000     8.822508    10.05503
-------------------------------------------------------------------------------------

. 
. ** Effect of ss_benefit100dollar on the log buy price
. reg logbuyprice  any_complexity consequence cognix_pca $exp_controls_linben $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(30, 4029)       =      10.16
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0630
                                                Root MSE          =      2.062

-------------------------------------------------------------------------------------
                    |               Robust
        logbuyprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------------+----------------------------------------------------------------
     any_complexity |  -.1396549   .0681654    -2.05   0.041    -.2732968   -.0060131
        consequence |   .1379157   .0649372     2.12   0.034      .010603    .2652285
         cognix_pca |   .1021603   .0457333     2.23   0.026     .0124978    .1918228
         sell_first |   .7729632   .0654481    11.81   0.000     .6446487    .9012777
    ls_startvalue_2 |   .2314519   .0794133     2.91   0.004     .0757579    .3871459
    ls_startvalue_3 |   .4756209   .0788994     6.03   0.000     .3209345    .6303073
           ls_first |  -.0656398   .0648706    -1.01   0.312    -.1928221    .0615426
ss_benefit100dollar |   -.025456   .0073446    -3.47   0.001    -.0398555   -.0110566
    vignette_name_2 |  -.0985818   .0886618    -1.11   0.266    -.2724079    .0752443
    vignette_name_3 |    .109945   .0916673     1.20   0.230    -.0697737    .2896637
    vignette_name_4 |   .1486637    .089143     1.67   0.095    -.0261059    .3234334
                age |  -.0345677   .0129543    -2.67   0.008    -.0599653   -.0091702
              agesq |   .0233031   .0129094     1.81   0.071    -.0020063    .0486126
             female |  -.1607905   .0689888    -2.33   0.020    -.2960467   -.0255343
            married |  -.1027367   .0815663    -1.26   0.208    -.2626517    .0571783
            nhblack |  -.0992579   .1480772    -0.67   0.503     -.389571    .1910552
            nhother |  -.0853021   .1287176    -0.66   0.508    -.3376598    .1670556
           hispanic |  -.0901746   .1331699    -0.68   0.498    -.3512612     .170912
         ed_dropout |   .1203119     .18184     0.66   0.508    -.2361951     .476819
         ed_hschool |   .0380501   .0992798     0.38   0.702    -.1565932    .2326934
         ed_college |   .0986563   .0866301     1.14   0.255    -.0711865    .2684991
         ed_graduat |   .2241959   .0995157     2.25   0.024       .02909    .4193017
         hinc_25_50 |   -.141823   .1232428    -1.15   0.250     -.383447    .0998009
         hinc_50_75 |   -.054672   .1217464    -0.45   0.653    -.2933623    .1840182
        hinc_75_100 |  -.0192808   .1316114    -0.15   0.884     -.277312    .2387504
         hinc_ge100 |  -.0484476   .1116401    -0.43   0.664     -.267324    .1704288
            hhsiz_2 |   -.014802   .0998821    -0.15   0.882    -.2106262    .1810222
            hhsiz_3 |  -.0236338   .1348086    -0.18   0.861    -.2879332    .2406657
           hhsiz_4p |  -.1975329   .1521538    -1.30   0.194    -.4958385    .1007726
            anykids |   .1059898   .1141864     0.93   0.353    -.1178787    .3298584
              _cons |   9.681941   .3633155    26.65   0.000     8.969641    10.39424
-------------------------------------------------------------------------------------

. 
. 
. 
. ** Text Claim 12 - Section 3.3  
. ** ---------------------------
. ** "To alleviate concerns about multiple hypotheses testing, we also test whether 
. ** our two key experimental manipulations, the consequence message and the complexity
. ** treatment, are jointly zero: we reject this hypothesis with a p-value of 0.0106. 
. ** The p-value becomes 0.0256 if we do not pool the complexity treatment, i.e., when 
. ** we test that the consequence message, the wide-age-range complexity treatment, and
. ** the extra-information complexity treatment are jointly zero. If we include all the 
. ** secondary experimental manipulations in the joint test, we can reject that all 
. ** treatment effects are jointly zero with a p-value of 0.0098 when the complexity 
. ** treatments are pooled and with a p-value of 0.0148 when the complexity treatments
. ** are separated out."
. 
. ** -------- Pooling Complexity Treatments ---------
. ** Re-run regression from table 3 for column 1 so that testparm command can be run
. reg logspread    any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      24.04
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1568
                                                Root MSE          =     1.9603

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1314293   .0648348     2.03   0.043     .0043173    .2585414
    consequence |  -.1405306   .0616294    -2.28   0.023    -.2613583   -.0197029
     cognix_pca |  -.7876547   .0427562   -18.42   0.000    -.8714804   -.7038289
     sell_first |   .1657743   .0616662     2.69   0.007     .0448745    .2866741
ls_startvalue_2 |   .0632191   .0761058     0.83   0.406    -.0859904    .2124286
ls_startvalue_3 |  -.0021141   .0749548    -0.03   0.978    -.1490669    .1448388
       ls_first |   .0294388   .0616149     0.48   0.633    -.0913606    .1502381
   ss_benefit_2 |    .112784   .0870307     1.30   0.195    -.0578444    .2834124
   ss_benefit_3 |   .0573012   .0842277     0.68   0.496    -.1078317     .222434
   ss_benefit_4 |   .1669049   .0868148     1.92   0.055    -.0033001    .3371099
vignette_name_2 |   .1140634      .0856     1.33   0.183    -.0537599    .2818867
vignette_name_3 |   .0882214   .0877756     1.01   0.315    -.0838674    .2603102
vignette_name_4 |  -.0109552   .0850845    -0.13   0.898    -.1777678    .1558574
            age |   .0247686   .0129824     1.91   0.056    -.0006842    .0502213
          agesq |  -.0147106   .0129063    -1.14   0.254     -.040014    .0105928
         female |   .0854357   .0663126     1.29   0.198    -.0445738    .2154451
        married |   .0965551   .0759943     1.27   0.204    -.0524358    .2455459
        nhblack |   .0281964   .1418699     0.20   0.842    -.2499472      .30634
        nhother |   .0481969   .1215336     0.40   0.692    -.1900762      .28647
       hispanic |   .0810356   .1251889     0.65   0.517    -.1644039    .3264752
     ed_dropout |  -.0573967   .1777837    -0.32   0.747     -.405951    .2911577
     ed_hschool |   .0334476   .0930188     0.36   0.719    -.1489207    .2158159
     ed_college |   .0076616   .0838392     0.09   0.927    -.1567096    .1720329
     ed_graduat |   .0763065   .0954393     0.80   0.424    -.1108073    .2634204
     hinc_25_50 |   .1019141   .1167251     0.87   0.383    -.1269317      .33076
     hinc_50_75 |  -.1660551    .115919    -1.43   0.152    -.3933204    .0612103
    hinc_75_100 |  -.0550164   .1299977    -0.42   0.672    -.3098838     .199851
     hinc_ge100 |  -.2568218   .1099252    -2.34   0.020    -.4723359   -.0413077
        hhsiz_2 |  -.0247458   .0954113    -0.26   0.795    -.2118047    .1623132
        hhsiz_3 |   .1468662   .1305622     1.12   0.261    -.1091079    .4028402
       hhsiz_4p |   .1815883   .1452072     1.25   0.211    -.1030981    .4662747
        anykids |  -.1769094   .1055735    -1.68   0.094    -.3838918    .0300729
          _cons |   1.106907   .3419853     3.24   0.001     .4364261    1.777387
---------------------------------------------------------------------------------

. 
. ** Joint test of key treatments
. testparm any_complexity consequence

 ( 1)  any_complexity = 0
 ( 2)  consequence = 0

       F(  2,  4027) =    4.55
            Prob > F =    0.0106

. 
. ** Joint test of all treatments including secondary ones (even though the secondary ones were not expected to be significant)
. testparm any_complexity consequence $exp_controls

 ( 1)  any_complexity = 0
 ( 2)  consequence = 0
 ( 3)  sell_first = 0
 ( 4)  ls_startvalue_2 = 0
 ( 5)  ls_startvalue_3 = 0
 ( 6)  ls_first = 0
 ( 7)  ss_benefit_2 = 0
 ( 8)  ss_benefit_3 = 0
 ( 9)  ss_benefit_4 = 0
 (10)  vignette_name_2 = 0
 (11)  vignette_name_3 = 0
 (12)  vignette_name_4 = 0

       F( 12,  4027) =    2.19
            Prob > F =    0.0098

. 
. ** -------- Not Pooling Complexity Treatments -------------
. ** Re-run regression from APX table A06 for column 1 so that testparm command can be run on this data
. reg logspread complexity_2 complexity_3 consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(33, 4026)       =      23.33
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1568
                                                Root MSE          =     1.9605

---------------------------------------------------------------------------------
                |               Robust
      logspread |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
   complexity_2 |   .1494204   .0763919     1.96   0.051      -.00035    .2991908
   complexity_3 |   .1139917   .0745133     1.53   0.126    -.0320957     .260079
    consequence |   -.140479   .0616361    -2.28   0.023    -.2613199    -.019638
     cognix_pca |  -.7877918   .0427462   -18.43   0.000    -.8715979   -.7039856
     sell_first |   .1654395    .061665     2.68   0.007      .044542    .2863371
ls_startvalue_2 |   .0636204   .0761229     0.84   0.403    -.0856227    .2128635
ls_startvalue_3 |   -.001235   .0749698    -0.02   0.987    -.1482173    .1457474
       ls_first |   .0302774   .0616314     0.49   0.623    -.0905543     .151109
   ss_benefit_2 |   .1121536   .0870487     1.29   0.198    -.0585101    .2828173
   ss_benefit_3 |   .0567269   .0843089     0.67   0.501    -.1085651    .2220189
   ss_benefit_4 |   .1670934   .0868058     1.92   0.054     -.003094    .3372808
vignette_name_2 |    .114343   .0856407     1.34   0.182    -.0535602    .2822462
vignette_name_3 |    .089812   .0880426     1.02   0.308    -.0828002    .2624241
vignette_name_4 |  -.0098805   .0852609    -0.12   0.908     -.177039     .157278
            age |   .0247629   .0129922     1.91   0.057     -.000709    .0502348
          agesq |  -.0146936   .0129151    -1.14   0.255    -.0400144    .0106272
         female |   .0853478    .066319     1.29   0.198    -.0446742    .2153698
        married |   .0959846   .0759945     1.26   0.207    -.0530068     .244976
        nhblack |    .028033   .1419029     0.20   0.843    -.2501753    .3062412
        nhother |   .0483704   .1215531     0.40   0.691     -.189941    .2866819
       hispanic |    .081131   .1251898     0.65   0.517    -.1643104    .3265723
     ed_dropout |  -.0585211   .1778733    -0.33   0.742    -.4072512     .290209
     ed_hschool |   .0338814   .0930291     0.36   0.716    -.1485072      .21627
     ed_college |   .0071948   .0838438     0.09   0.932    -.1571855    .1715751
     ed_graduat |   .0766582   .0954304     0.80   0.422    -.1104382    .2637546
     hinc_25_50 |   .1020894   .1167399     0.87   0.382    -.1267854    .3309642
     hinc_50_75 |  -.1670493   .1159924    -1.44   0.150    -.3944586      .06036
    hinc_75_100 |  -.0555349   .1300357    -0.43   0.669    -.3104769    .1994071
     hinc_ge100 |  -.2576225   .1099666    -2.34   0.019    -.4732179   -.0420271
        hhsiz_2 |  -.0243049   .0954436    -0.25   0.799    -.2114271    .1628173
        hhsiz_3 |   .1457747    .130605     1.12   0.264    -.1102833    .4018328
       hhsiz_4p |   .1798826   .1453796     1.24   0.216    -.1051418    .4649071
        anykids |  -.1754011   .1056487    -1.66   0.097    -.3825309    .0317288
          _cons |   1.106359   .3422105     3.23   0.001     .4354372    1.777281
---------------------------------------------------------------------------------

. 
. ** Joint test of key treatments
. testparm complexity_2 complexity_3 consequence

 ( 1)  complexity_2 = 0
 ( 2)  complexity_3 = 0
 ( 3)  consequence = 0

       F(  3,  4026) =    3.10
            Prob > F =    0.0256

. 
. ** Joint test of all treatments including secondary ones (even though the secondary ones were not expected to be significant)
. testparm complexity_2 complexity_3 consequence $exp_controls

 ( 1)  complexity_2 = 0
 ( 2)  complexity_3 = 0
 ( 3)  consequence = 0
 ( 4)  sell_first = 0
 ( 5)  ls_startvalue_2 = 0
 ( 6)  ls_startvalue_3 = 0
 ( 7)  ls_first = 0
 ( 8)  ss_benefit_2 = 0
 ( 9)  ss_benefit_3 = 0
 (10)  ss_benefit_4 = 0
 (11)  vignette_name_2 = 0
 (12)  vignette_name_3 = 0
 (13)  vignette_name_4 = 0

       F( 13,  4026) =    2.04
            Prob > F =    0.0148

. 
. 
. 
. 
. 
. ** Text Claim 13 - Section 3.3  
. ** ---------------------------
. ** "What would annuity valuations be if we had an intervention sufficiently powerful
. ** to cause the mean log sell price and the mean log buy price to be equal (so no 
. ** deviation from rationality at the mean)? We can get a rough answer to this question 
. ** by extrapolating the effects of each of our two main experimental interventions. 
. ** The mean log difference between sell and buy price is 1.01 (see Figure 2), and the
. ** consequence message moves log sell and buy price closer by 0.122 (=0.133-0.011, see 
. ** columns 2 and 3 of Table 3). Thus, a treatment about 8 ≈ 1.01/0.122 times more powerful 
. ** than our current consequence message would close the gap between the mean log sell and buy 
. ** price. At that level of treatment, the median sell and buy price would be predicted to
. ** be about $17,000. Similarly, we can extrapolate the complexity treatment, in the direction
. ** of making the problem less complex, such that the sell and buy price coincide. This would 
. ** require reducing complexity by about 5 times the amount of complexity added by our 
. ** complexity treatment. The resulting sell and buy price would then be predicted to be
. ** about $12,000."
. 
. 
. ** 1. Predict buy and sell values at CONSEQUENCE level that causes log sell and by prices to be equal
. ** Run auxiliary regression of the log buy - sell difference (not in absolute values):
. 
. ** Generate mean_logsellbuydiff global 
. sum logsellbuydiff if basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
logsellbuy~f |      4,060    1.009391    2.899622  -8.006368   8.006368

. global mean_logsellbuydiff=r(mean)

. 
. reg logsellbuydiff    any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =       6.91
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0509
                                                Root MSE          =     2.8361

---------------------------------------------------------------------------------
                |               Robust
 logsellbuydiff |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1864048   .0936656     1.99   0.047     .0027684    .3700412
    consequence |  -.1228253   .0891115    -1.38   0.168    -.2975332    .0518826
     cognix_pca |  -.2865081   .0640007    -4.48   0.000    -.4119849   -.1610313
     sell_first |  -.8194922   .0894449    -9.16   0.000    -.9948538   -.6441307
ls_startvalue_2 |   .0037863   .1097893     0.03   0.972    -.2114615    .2190341
ls_startvalue_3 |   .0078537    .108549     0.07   0.942    -.2049625    .2206699
       ls_first |   .0213265    .089205     0.24   0.811    -.1535648    .1962177
   ss_benefit_2 |   .4685725   .1254709     3.73   0.000     .2225801     .714565
   ss_benefit_3 |   .3874067   .1218385     3.18   0.001     .1485358    .6262775
   ss_benefit_4 |   .2349896   .1288953     1.82   0.068    -.0177164    .4876956
vignette_name_2 |   .0698161   .1230676     0.57   0.571    -.1714645    .3110967
vignette_name_3 |  -.2111705   .1256923    -1.68   0.093     -.457597     .035256
vignette_name_4 |    -.22739   .1226673    -1.85   0.064    -.4678858    .0131058
            age |   .0360055   .0189191     1.90   0.057    -.0010865    .0730975
          agesq |  -.0177112   .0186178    -0.95   0.342    -.0542125      .01879
         female |   .0841035   .0944397     0.89   0.373    -.1010505    .2692575
        married |   .0968469    .112514     0.86   0.389    -.1237428    .3174366
        nhblack |   .0285683   .2155744     0.13   0.895    -.3940767    .4512134
        nhother |    .031412   .1786079     0.18   0.860    -.3187583    .3815822
       hispanic |   .0026086   .1930729     0.01   0.989     -.375921    .3811383
     ed_dropout |   .0024084   .2665627     0.01   0.993     -.520202    .5250189
     ed_hschool |   .0559901   .1380344     0.41   0.685    -.2146338    .3266139
     ed_college |  -.0806501   .1183724    -0.68   0.496    -.3127255    .1514253
     ed_graduat |  -.0091614   .1321123    -0.07   0.945    -.2681746    .2498517
     hinc_25_50 |    .182225   .1772423     1.03   0.304    -.1652679     .529718
     hinc_50_75 |  -.0329127   .1712842    -0.19   0.848    -.3687245    .3028992
    hinc_75_100 |  -.0937965   .1887949    -0.50   0.619     -.463939    .2763461
     hinc_ge100 |   -.057574   .1607084    -0.36   0.720    -.3726513    .2575034
        hhsiz_2 |   .0487451   .1400094     0.35   0.728    -.2257508    .3232411
        hhsiz_3 |   .2774669     .19127     1.45   0.147    -.0975282     .652462
       hhsiz_4p |   .3850625   .2150736     1.79   0.073    -.0366007    .8067257
        anykids |  -.3556887   .1571141    -2.26   0.024    -.6637192   -.0476581
          _cons |  -.2631023   .5124055    -0.51   0.608    -1.267701    .7414959
---------------------------------------------------------------------------------

. sum consequence if basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
 consequence |      4,060    .5078818    .4999995          0          1

. global required_consequence = r(mean) -  $mean_logsellbuydiff/_b[consequence] 

. 
. di "Level of consequence where log buy and sell prices are equal: " $required_consequence
Level of consequence where log buy and sell prices are equal: 8.7259811

. 
. ** Now predict logbuyprice at the "required" level of consequence
. reg logbuyprice   any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      10.12
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0672
                                                Root MSE          =     2.0579

---------------------------------------------------------------------------------
                |               Robust
    logbuyprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |  -.1367219    .068086    -2.01   0.045    -.2702081   -.0032356
    consequence |   .1334814   .0648055     2.06   0.039     .0064268    .2605359
     cognix_pca |   .0981797   .0456418     2.15   0.032     .0086965    .1876629
     sell_first |   .7769918   .0653356    11.89   0.000     .6488978    .9050858
ls_startvalue_2 |   .2355473   .0792337     2.97   0.003     .0802054    .3908893
ls_startvalue_3 |   .4764367   .0788874     6.04   0.000     .3217737    .6310996
       ls_first |  -.0649965   .0647657    -1.00   0.316    -.1919731    .0619801
   ss_benefit_2 |  -.4584842   .0925164    -4.96   0.000    -.6398674    -.277101
   ss_benefit_3 |  -.3934582   .0906761    -4.34   0.000    -.5712335   -.2156829
   ss_benefit_4 |  -.3534632    .092986    -3.80   0.000    -.5357672   -.1711591
vignette_name_2 |  -.0983106   .0886346    -1.11   0.267    -.2720835    .0754624
vignette_name_3 |   .1140095   .0913234     1.25   0.212    -.0650349    .2930539
vignette_name_4 |   .1462166   .0890904     1.64   0.101    -.0284498     .320883
            age |   -.034668   .0129412    -2.68   0.007    -.0600399   -.0092961
          agesq |    .023434   .0129005     1.82   0.069    -.0018581    .0487262
         female |  -.1595815   .0687841    -2.32   0.020    -.2944364   -.0247265
        married |  -.1037356    .081465    -1.27   0.203    -.2634522    .0559809
        nhblack |  -.1158509   .1477744    -0.78   0.433    -.4055704    .1738686
        nhother |   -.087201   .1283783    -0.68   0.497    -.3388934    .1644915
       hispanic |  -.0968947   .1329446    -0.73   0.466    -.3575396    .1637502
     ed_dropout |    .135633   .1818374     0.75   0.456     -.220869    .4921349
     ed_hschool |    .047632   .0989367     0.48   0.630    -.1463387    .2416028
     ed_college |   .1000541   .0864912     1.16   0.247    -.0695165    .2696246
     ed_graduat |   .2327128   .0995416     2.34   0.019     .0375562    .4278693
     hinc_25_50 |  -.1358911   .1228556    -1.11   0.269    -.3767561     .104974
     hinc_50_75 |  -.0372692    .121471    -0.31   0.759    -.2754195    .2008811
    hinc_75_100 |  -.0104659   .1316783    -0.08   0.937    -.2686282    .2476965
     hinc_ge100 |  -.0407582   .1112642    -0.37   0.714    -.2588977    .1773812
        hhsiz_2 |  -.0070515   .0996338    -0.07   0.944    -.2023889    .1882858
        hhsiz_3 |  -.0250649   .1343336    -0.19   0.852     -.288433    .2383032
       hhsiz_4p |  -.2070519   .1511576    -1.37   0.171    -.5034044    .0893007
        anykids |   .1168671   .1135367     1.03   0.303    -.1057277     .339462
          _cons |   9.608916   .3506399    27.40   0.000     8.921467    10.29636
---------------------------------------------------------------------------------

. gen logbuyprice_reqconsequence = logbuyprice + _b[consequence]*($required_consequence - consequence) if basesample
(536 missing values generated)

. gen    buyprice_reqconsequence = exp(logbuyprice_reqconsequence)
(536 missing values generated)

. 
. sum logbuyprice                if basesample, d

                           Log Buy
-------------------------------------------------------------
      Percentiles      Smallest
 1%     5.521461       5.521461
 5%     5.521461       5.521461
10%     5.521461       5.521461       Obs               4,060
25%     7.130899       5.521461       Sum of Wgt.       4,060

50%     8.678461                      Mean           8.668611
                        Largest       Std. Dev.      2.122283
75%     9.695848       13.52783
90%     11.22524       13.52783       Variance       4.504085
95%     13.52783       13.52783       Skewness       .5208394
99%     13.52783       13.52783       Kurtosis       3.003292

. sum logbuyprice_reqconsequence if basesample, d

                 logbuyprice_reqconsequence
-------------------------------------------------------------
      Percentiles      Smallest
 1%     6.552735       6.552735
 5%     6.552735       6.552735
10%     6.686217       6.552735       Obs               4,060
25%     8.295654       6.552735       Sum of Wgt.       4,060

50%     9.709736                      Mean           9.765574
                        Largest       Std. Dev.      2.121296
75%      10.8606       14.69258
90%     12.38168       14.69258       Variance       4.499896
95%      14.5591       14.69258       Skewness       .5269284
99%     14.69258       14.69258       Kurtosis       3.010512

. 
. di "The exp of log buy price that equals log sell price is: " exp(r(mean))
The exp of log buy price that equals log sell price is: 17423.474

. 
. sum midbuy                  if basesample, d

                    Buy price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          250            250
10%          250            250       Obs               4,060
25%         1250            250       Sum of Wgt.       4,060

50%         5875                      Mean           67327.09
                        Largest       Std. Dev.      189302.6
75%        16250         750000
90%        75000         750000       Variance       3.58e+10
95%       750000         750000       Skewness       3.072675
99%       750000         750000       Kurtosis       10.70886

. sum buyprice_reqconsequence if basesample, d

                   buyprice_reqconsequence
-------------------------------------------------------------
      Percentiles      Smallest
 1%     701.1595       701.1595
 5%     701.1595       701.1595
10%     801.2851       701.1595       Obs               4,060
25%     4006.424       701.1595       Sum of Wgt.       4,060

50%     16477.25                      Mean           202372.5
                        Largest       Std. Dev.      570485.5
75%     52083.54        2403855
90%     238394.3        2403855       Variance       3.25e+11
95%      2103479        2403855       Skewness       3.095745
99%      2403855        2403855       Kurtosis       10.91355

. 
. ** Now predict logsellprice at the "required" level of consequence
. reg logsellprice   any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =       4.60
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0350
                                                Root MSE          =     1.7382

---------------------------------------------------------------------------------
                |               Robust
   logsellprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |    .049683   .0574995     0.86   0.388    -.0630479    .1624138
    consequence |    .010656   .0547079     0.19   0.846    -.0966018    .1179138
     cognix_pca |  -.1883284   .0381627    -4.93   0.000    -.2631484   -.1135084
     sell_first |  -.0425004   .0545079    -0.78   0.436     -.149366    .0643652
ls_startvalue_2 |   .2393336   .0665975     3.59   0.000     .1087658    .3699015
ls_startvalue_3 |   .4842904   .0675399     7.17   0.000     .3518748    .6167059
       ls_first |    -.04367    .054771    -0.80   0.425    -.1510515    .0637114
   ss_benefit_2 |   .0100883   .0754805     0.13   0.894    -.1378952    .1580718
   ss_benefit_3 |  -.0060516   .0739202    -0.08   0.935    -.1509761     .138873
   ss_benefit_4 |  -.1184736   .0798293    -1.48   0.138    -.2749832    .0380361
vignette_name_2 |  -.0284945   .0760599    -0.37   0.708     -.177614    .1206251
vignette_name_3 |   -.097161    .076411    -1.27   0.204    -.2469688    .0526468
vignette_name_4 |  -.0811734   .0760992    -1.07   0.286      -.23037    .0680232
            age |   .0013375   .0108285     0.12   0.902    -.0198924    .0225673
          agesq |   .0057228   .0103572     0.55   0.581    -.0145831    .0260287
         female |   -.075478   .0584121    -1.29   0.196    -.1899981    .0390421
        married |  -.0068887   .0694248    -0.10   0.921    -.1429997    .1292223
        nhblack |  -.0872826   .1342889    -0.65   0.516     -.350563    .1759979
        nhother |   -.055789   .1074913    -0.52   0.604    -.2665314    .1549534
       hispanic |  -.0942861   .1215208    -0.78   0.438    -.3325341     .143962
     ed_dropout |   .1380414   .1609612     0.86   0.391    -.1775316    .4536144
     ed_hschool |   .1036221   .0854708     1.21   0.225    -.0639479    .2711921
     ed_college |    .019404   .0730489     0.27   0.791    -.1238122    .1626202
     ed_graduat |   .2235513   .0765962     2.92   0.004     .0733803    .3737223
     hinc_25_50 |    .046334   .1082473     0.43   0.669    -.1658907    .2585586
     hinc_50_75 |  -.0701819    .107471    -0.65   0.514    -.2808846    .1405207
    hinc_75_100 |  -.1042624   .1189632    -0.88   0.381    -.3374961    .1289714
     hinc_ge100 |  -.0983322   .1002532    -0.98   0.327     -.294884    .0982196
        hhsiz_2 |   .0416936   .0836537     0.50   0.618    -.1223138     .205701
        hhsiz_3 |    .252402   .1134832     2.22   0.026     .0299121    .4748919
       hhsiz_4p |   .1780106   .1346359     1.32   0.186    -.0859502    .4419714
        anykids |  -.2388215   .1008865    -2.37   0.018    -.4366148   -.0410283
          _cons |   9.345813   .3058455    30.56   0.000     8.746187     9.94544
---------------------------------------------------------------------------------

. gen logsellprice_reqconsequence = logsellprice + _b[consequence]*($required_consequence - consequence) if basesample
(536 missing values generated)

. gen    sellprice_reqconsequence = exp(logsellprice_reqconsequence)
(536 missing values generated)

. 
. sum logsellprice                if basesample, d

                          Log Sell
-------------------------------------------------------------
      Percentiles      Smallest
 1%     5.521461       5.521461
 5%     6.214608       5.521461
10%     7.467371       5.521461       Obs               4,060
25%     8.798606       5.521461       Sum of Wgt.       4,060

50%     9.695848                      Mean           9.678001
                        Largest       Std. Dev.      1.762509
75%       10.389       13.52783
90%     12.07254       13.52783       Variance       3.106437
95%     13.30468       13.52783       Skewness       .1356991
99%     13.52783       13.52783       Kurtosis       3.474863

. sum logsellprice_reqconsequence if basesample, d

                 logsellprice_reqconsequence
-------------------------------------------------------------
      Percentiles      Smallest
 1%     5.603789       5.603789
 5%     6.307592       5.603789
10%     7.560356       5.603789       Obs               4,060
25%     8.880934       5.603789       Sum of Wgt.       4,060

50%     9.778176                      Mean           9.765574
                        Largest       Std. Dev.      1.762506
75%     10.47665       13.62081
90%     12.15487       13.62081       Variance       3.106426
95%     13.39767       13.62081       Skewness       .1362284
99%     13.62081       13.62081       Kurtosis       3.474761

. 
. di "The exp of log sell price that equals log buy price is: " exp(r(mean))
The exp of log sell price that equals log buy price is: 17423.475

. 
. sum midsell                  if basesample, d

                    Sell price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          500            250
10%         1750            250       Obs               4,060
25%         6625            250       Sum of Wgt.       4,060

50%        16250                      Mean            79280.7
                        Largest       Std. Dev.      184016.4
75%        32500         750000
90%       175000         750000       Variance       3.39e+10
95%       600000         750000       Skewness       2.947456
99%       750000         750000       Kurtosis       10.27378

. sum sellprice_reqconsequence if basesample, d

                  sellprice_reqconsequence
-------------------------------------------------------------
      Percentiles      Smallest
 1%     271.4531       271.4531
 5%     548.7222       271.4531
10%     1920.528       271.4531       Obs               4,060
25%     7193.504       271.4531       Sum of Wgt.       4,060

50%     17644.45                      Mean           86559.07
                        Largest       Std. Dev.      200938.9
75%     35477.92       823083.9
90%     190017.2       823083.9       Variance       4.04e+10
95%     658467.1       823083.9       Skewness       2.947762
99%     823083.9       823083.9       Kurtosis       10.27535

. 
. 
. ** 2. Predict buy and sell values at COMPLEXITY level that causes log sell and by prices to be equal
. ** Run auxiliary regression of the log buy - sell difference (not in absolute values):
. reg logsellbuydiff    any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =       6.91
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0509
                                                Root MSE          =     2.8361

---------------------------------------------------------------------------------
                |               Robust
 logsellbuydiff |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |   .1864048   .0936656     1.99   0.047     .0027684    .3700412
    consequence |  -.1228253   .0891115    -1.38   0.168    -.2975332    .0518826
     cognix_pca |  -.2865081   .0640007    -4.48   0.000    -.4119849   -.1610313
     sell_first |  -.8194922   .0894449    -9.16   0.000    -.9948538   -.6441307
ls_startvalue_2 |   .0037863   .1097893     0.03   0.972    -.2114615    .2190341
ls_startvalue_3 |   .0078537    .108549     0.07   0.942    -.2049625    .2206699
       ls_first |   .0213265    .089205     0.24   0.811    -.1535648    .1962177
   ss_benefit_2 |   .4685725   .1254709     3.73   0.000     .2225801     .714565
   ss_benefit_3 |   .3874067   .1218385     3.18   0.001     .1485358    .6262775
   ss_benefit_4 |   .2349896   .1288953     1.82   0.068    -.0177164    .4876956
vignette_name_2 |   .0698161   .1230676     0.57   0.571    -.1714645    .3110967
vignette_name_3 |  -.2111705   .1256923    -1.68   0.093     -.457597     .035256
vignette_name_4 |    -.22739   .1226673    -1.85   0.064    -.4678858    .0131058
            age |   .0360055   .0189191     1.90   0.057    -.0010865    .0730975
          agesq |  -.0177112   .0186178    -0.95   0.342    -.0542125      .01879
         female |   .0841035   .0944397     0.89   0.373    -.1010505    .2692575
        married |   .0968469    .112514     0.86   0.389    -.1237428    .3174366
        nhblack |   .0285683   .2155744     0.13   0.895    -.3940767    .4512134
        nhother |    .031412   .1786079     0.18   0.860    -.3187583    .3815822
       hispanic |   .0026086   .1930729     0.01   0.989     -.375921    .3811383
     ed_dropout |   .0024084   .2665627     0.01   0.993     -.520202    .5250189
     ed_hschool |   .0559901   .1380344     0.41   0.685    -.2146338    .3266139
     ed_college |  -.0806501   .1183724    -0.68   0.496    -.3127255    .1514253
     ed_graduat |  -.0091614   .1321123    -0.07   0.945    -.2681746    .2498517
     hinc_25_50 |    .182225   .1772423     1.03   0.304    -.1652679     .529718
     hinc_50_75 |  -.0329127   .1712842    -0.19   0.848    -.3687245    .3028992
    hinc_75_100 |  -.0937965   .1887949    -0.50   0.619     -.463939    .2763461
     hinc_ge100 |   -.057574   .1607084    -0.36   0.720    -.3726513    .2575034
        hhsiz_2 |   .0487451   .1400094     0.35   0.728    -.2257508    .3232411
        hhsiz_3 |   .2774669     .19127     1.45   0.147    -.0975282     .652462
       hhsiz_4p |   .3850625   .2150736     1.79   0.073    -.0366007    .8067257
        anykids |  -.3556887   .1571141    -2.26   0.024    -.6637192   -.0476581
          _cons |  -.2631023   .5124055    -0.51   0.608    -1.267701    .7414959
---------------------------------------------------------------------------------

. sum any_complexity if basesample

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
any_comple~y |      4,060    .6529557    .4760886          0          1

. global required_complexity = r(mean) -  $mean_logsellbuydiff/_b[any_complexity]

. di "Level of complexity where log buy and sell prices are equal: " $required_complexity
Level of complexity where log buy and sell prices are equal: -4.7620901

. 
. ** Now predict logbuyprice at the "required" level of complexity
. reg logbuyprice   any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =      10.12
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0672
                                                Root MSE          =     2.0579

---------------------------------------------------------------------------------
                |               Robust
    logbuyprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |  -.1367219    .068086    -2.01   0.045    -.2702081   -.0032356
    consequence |   .1334814   .0648055     2.06   0.039     .0064268    .2605359
     cognix_pca |   .0981797   .0456418     2.15   0.032     .0086965    .1876629
     sell_first |   .7769918   .0653356    11.89   0.000     .6488978    .9050858
ls_startvalue_2 |   .2355473   .0792337     2.97   0.003     .0802054    .3908893
ls_startvalue_3 |   .4764367   .0788874     6.04   0.000     .3217737    .6310996
       ls_first |  -.0649965   .0647657    -1.00   0.316    -.1919731    .0619801
   ss_benefit_2 |  -.4584842   .0925164    -4.96   0.000    -.6398674    -.277101
   ss_benefit_3 |  -.3934582   .0906761    -4.34   0.000    -.5712335   -.2156829
   ss_benefit_4 |  -.3534632    .092986    -3.80   0.000    -.5357672   -.1711591
vignette_name_2 |  -.0983106   .0886346    -1.11   0.267    -.2720835    .0754624
vignette_name_3 |   .1140095   .0913234     1.25   0.212    -.0650349    .2930539
vignette_name_4 |   .1462166   .0890904     1.64   0.101    -.0284498     .320883
            age |   -.034668   .0129412    -2.68   0.007    -.0600399   -.0092961
          agesq |    .023434   .0129005     1.82   0.069    -.0018581    .0487262
         female |  -.1595815   .0687841    -2.32   0.020    -.2944364   -.0247265
        married |  -.1037356    .081465    -1.27   0.203    -.2634522    .0559809
        nhblack |  -.1158509   .1477744    -0.78   0.433    -.4055704    .1738686
        nhother |   -.087201   .1283783    -0.68   0.497    -.3388934    .1644915
       hispanic |  -.0968947   .1329446    -0.73   0.466    -.3575396    .1637502
     ed_dropout |    .135633   .1818374     0.75   0.456     -.220869    .4921349
     ed_hschool |    .047632   .0989367     0.48   0.630    -.1463387    .2416028
     ed_college |   .1000541   .0864912     1.16   0.247    -.0695165    .2696246
     ed_graduat |   .2327128   .0995416     2.34   0.019     .0375562    .4278693
     hinc_25_50 |  -.1358911   .1228556    -1.11   0.269    -.3767561     .104974
     hinc_50_75 |  -.0372692    .121471    -0.31   0.759    -.2754195    .2008811
    hinc_75_100 |  -.0104659   .1316783    -0.08   0.937    -.2686282    .2476965
     hinc_ge100 |  -.0407582   .1112642    -0.37   0.714    -.2588977    .1773812
        hhsiz_2 |  -.0070515   .0996338    -0.07   0.944    -.2023889    .1882858
        hhsiz_3 |  -.0250649   .1343336    -0.19   0.852     -.288433    .2383032
       hhsiz_4p |  -.2070519   .1511576    -1.37   0.171    -.5034044    .0893007
        anykids |   .1168671   .1135367     1.03   0.303    -.1057277     .339462
          _cons |   9.608916   .3506399    27.40   0.000     8.921467    10.29636
---------------------------------------------------------------------------------

. gen logbuyprice_reqcomplex = logbuyprice + _b[any_complexity]*($required_complexity - any_complexity) if basesample
(536 missing values generated)

. gen    buyprice_reqcomplex = exp(logbuyprice_reqcomplex)
(536 missing values generated)

. 
. sum logbuyprice            if basesample, d

                           Log Buy
-------------------------------------------------------------
      Percentiles      Smallest
 1%     5.521461       5.521461
 5%     5.521461       5.521461
10%     5.521461       5.521461       Obs               4,060
25%     7.130899       5.521461       Sum of Wgt.       4,060

50%     8.678461                      Mean           8.668611
                        Largest       Std. Dev.      2.122283
75%     9.695848       13.52783
90%     11.22524       13.52783       Variance       4.504085
95%     13.52783       13.52783       Skewness       .5208394
99%     13.52783       13.52783       Kurtosis       3.003292

. sum logbuyprice_reqcomplex if basesample, d

                   logbuyprice_reqcomplex
-------------------------------------------------------------
      Percentiles      Smallest
 1%     6.172543       6.172543
 5%     6.309265       6.172543
10%     6.309265       6.172543       Obs               4,060
25%     7.918703       6.172543       Sum of Wgt.       4,060

50%     9.391418                      Mean           9.408966
                        Largest       Std. Dev.      2.121125
75%     10.48365       14.31563
90%     12.01305       14.31563       Variance       4.499172
95%     14.17891       14.31563       Skewness       .5199579
99%     14.31563       14.31563       Kurtosis       3.008619

. 
. di "The exp of log buy price that equals log sell price is: " exp(r(mean))
The exp of log buy price that equals log sell price is: 12197.249

. 
. sum midbuy              if basesample, d

                    Buy price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          250            250
10%          250            250       Obs               4,060
25%         1250            250       Sum of Wgt.       4,060

50%         5875                      Mean           67327.09
                        Largest       Std. Dev.      189302.6
75%        16250         750000
90%        75000         750000       Variance       3.58e+10
95%       750000         750000       Skewness       3.072675
99%       750000         750000       Kurtosis       10.70886

. sum buyprice_reqcomplex if basesample, d

                     buyprice_reqcomplex
-------------------------------------------------------------
      Percentiles      Smallest
 1%     479.4035       479.4035
 5%     549.6406       479.4035
10%     549.6406       479.4035       Obs               4,060
25%     2748.203       479.4035       Sum of Wgt.       4,060

50%     11985.09                      Mean             141088
                        Largest       Std. Dev.      397624.4
75%     35726.65        1648921
90%     164892.2        1648921       Variance       1.58e+11
95%      1438211        1648921       Skewness       3.094365
99%      1648921        1648921       Kurtosis       10.89525

. 
. ** Now predict logsellprice at the "required" level of complexity
. reg logsellprice   any_complexity consequence cognix_pca $exp_controls $demographics if basesample, robust

Linear regression                               Number of obs     =      4,060
                                                F(32, 4027)       =       4.60
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0350
                                                Root MSE          =     1.7382

---------------------------------------------------------------------------------
                |               Robust
   logsellprice |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
 any_complexity |    .049683   .0574995     0.86   0.388    -.0630479    .1624138
    consequence |    .010656   .0547079     0.19   0.846    -.0966018    .1179138
     cognix_pca |  -.1883284   .0381627    -4.93   0.000    -.2631484   -.1135084
     sell_first |  -.0425004   .0545079    -0.78   0.436     -.149366    .0643652
ls_startvalue_2 |   .2393336   .0665975     3.59   0.000     .1087658    .3699015
ls_startvalue_3 |   .4842904   .0675399     7.17   0.000     .3518748    .6167059
       ls_first |    -.04367    .054771    -0.80   0.425    -.1510515    .0637114
   ss_benefit_2 |   .0100883   .0754805     0.13   0.894    -.1378952    .1580718
   ss_benefit_3 |  -.0060516   .0739202    -0.08   0.935    -.1509761     .138873
   ss_benefit_4 |  -.1184736   .0798293    -1.48   0.138    -.2749832    .0380361
vignette_name_2 |  -.0284945   .0760599    -0.37   0.708     -.177614    .1206251
vignette_name_3 |   -.097161    .076411    -1.27   0.204    -.2469688    .0526468
vignette_name_4 |  -.0811734   .0760992    -1.07   0.286      -.23037    .0680232
            age |   .0013375   .0108285     0.12   0.902    -.0198924    .0225673
          agesq |   .0057228   .0103572     0.55   0.581    -.0145831    .0260287
         female |   -.075478   .0584121    -1.29   0.196    -.1899981    .0390421
        married |  -.0068887   .0694248    -0.10   0.921    -.1429997    .1292223
        nhblack |  -.0872826   .1342889    -0.65   0.516     -.350563    .1759979
        nhother |   -.055789   .1074913    -0.52   0.604    -.2665314    .1549534
       hispanic |  -.0942861   .1215208    -0.78   0.438    -.3325341     .143962
     ed_dropout |   .1380414   .1609612     0.86   0.391    -.1775316    .4536144
     ed_hschool |   .1036221   .0854708     1.21   0.225    -.0639479    .2711921
     ed_college |    .019404   .0730489     0.27   0.791    -.1238122    .1626202
     ed_graduat |   .2235513   .0765962     2.92   0.004     .0733803    .3737223
     hinc_25_50 |    .046334   .1082473     0.43   0.669    -.1658907    .2585586
     hinc_50_75 |  -.0701819    .107471    -0.65   0.514    -.2808846    .1405207
    hinc_75_100 |  -.1042624   .1189632    -0.88   0.381    -.3374961    .1289714
     hinc_ge100 |  -.0983322   .1002532    -0.98   0.327     -.294884    .0982196
        hhsiz_2 |   .0416936   .0836537     0.50   0.618    -.1223138     .205701
        hhsiz_3 |    .252402   .1134832     2.22   0.026     .0299121    .4748919
       hhsiz_4p |   .1780106   .1346359     1.32   0.186    -.0859502    .4419714
        anykids |  -.2388215   .1008865    -2.37   0.018    -.4366148   -.0410283
          _cons |   9.345813   .3058455    30.56   0.000     8.746187     9.94544
---------------------------------------------------------------------------------

. gen logsellprice_reqcomplex = logsellprice + _b[any_complexity]*($required_complexity - any_complexity) if basesample
(536 missing values generated)

. gen    sellprice_reqcomplex = exp(logsellprice_reqcomplex)
(536 missing values generated)

. 
. sum logsellprice            if basesample, d

                          Log Sell
-------------------------------------------------------------
      Percentiles      Smallest
 1%     5.521461       5.521461
 5%     6.214608       5.521461
10%     7.467371       5.521461       Obs               4,060
25%     8.798606       5.521461       Sum of Wgt.       4,060

50%     9.695848                      Mean           9.678001
                        Largest       Std. Dev.      1.762509
75%       10.389       13.52783
90%     12.07254       13.52783       Variance       3.106437
95%     13.30468       13.52783       Skewness       .1356991
99%     13.52783       13.52783       Kurtosis       3.474863

. sum logsellprice_reqcomplex if basesample, d

                   logsellprice_reqcomplex
-------------------------------------------------------------
      Percentiles      Smallest
 1%     5.235183       5.235183
 5%     5.978014       5.235183
10%     7.230776       5.235183       Obs               4,060
25%     8.512328       5.235183       Sum of Wgt.       4,060

50%     9.409571                      Mean           9.408966
                        Largest       Std. Dev.      1.762384
75%     10.11318       13.29123
90%     11.78626       13.29123       Variance       3.105997
95%     13.06809       13.29123       Skewness       .1351989
99%     13.29123       13.29123       Kurtosis        3.47512

. 
. di "The exp of log sell price that equals log buy price is: " exp(r(mean))
The exp of log sell price that equals log buy price is: 12197.249

. 
. sum midsell              if basesample, d

                    Sell price (midpoint)
-------------------------------------------------------------
      Percentiles      Smallest
 1%          250            250
 5%          500            250
10%         1750            250       Obs               4,060
25%         6625            250       Sum of Wgt.       4,060

50%        16250                      Mean            79280.7
                        Largest       Std. Dev.      184016.4
75%        32500         750000
90%       175000         750000       Variance       3.39e+10
95%       600000         750000       Skewness       2.947456
99%       750000         750000       Kurtosis       10.27378

. sum sellprice_reqcomplex if basesample, d

                    sellprice_reqcomplex
-------------------------------------------------------------
      Percentiles      Smallest
 1%     187.7635       187.7635
 5%     394.6556       187.7635
10%     1381.294       187.7635       Obs               4,060
25%     4975.734       187.7635       Sum of Wgt.       4,060

50%     12204.63                      Mean           60549.35
                        Largest       Std. Dev.      140528.9
75%     24665.98       591983.5
90%     131434.4       591983.5       Variance       1.97e+10
95%     473586.8       591983.5       Skewness       2.949347
99%     591983.5       591983.5       Kurtosis       10.29056

. 
. 
. 
. ** Text Claim 14 - Appendix "Discussion of Robustness" 
. ** ---------------------------------------------------
. 
. ** "This implies that there is an implicit topcode of $100,000 on the buy valuations,
. ** though respondents are permitted to give a buy recommendation at a price in excess
. ** of $100,000, and 9% of them do so."
. 
. ** For mentioning in the text: the fraction of buy valuations exceeding $100k
. gen topcodebuy = midbuy>100000 if basesample
(536 missing values generated)

. tab topcodebuy

 topcodebuy |      Freq.     Percent        Cum.
------------+-----------------------------------
          0 |      3,679       90.62       90.62
          1 |        381        9.38      100.00
------------+-----------------------------------
      Total |      4,060      100.00

.         
.         
. 
. log close
      name:  <unnamed>
       log:  /Users/erzoluttmer/Dropbox/AnnuityVignettes/VignFINAL_REStat/Analysis/annuityanalyses.log
  log type:  text
 closed on:  13 Dec 2019, 11:01:04
--------------------------------------------------------------------------------------------------------------------------------------------
